[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CV06AVKZ3G0MJWRWAJRFXF":3,"skill-same:01M7CV06AVKZ3G0MJWRWAJRFXF":295},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":13,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":54,"install":55,"resources":68},"01M7CV06AVKZ3G0MJWRWAJRFXF","unbroker",null,"Autonomously remove your info from data-broker sites.","dwikanugraha13\u002FProject-Anara","dwikanugraha13","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F63535036?v=4","01M7CTBN0DG8GDNP4QZ68W8M0A","01M7CTBN10Q015JDBYJS6AJRGN",0,"MIT","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: unbroker\ndescription: Autonomously remove your info from data-broker sites.\nversion: 1.0.0\nauthor: SHL0MS (github.com\u002FSHL0MS)\nlicense: MIT\nplatforms: [linux, macos, windows]\nprerequisites:\n  commands: [python]\nmetadata:\n  anara:\n    tags: [privacy, data-broker, opt-out, ccpa, gdpr, security, doxxing]\n    category: security\n    related_skills: [google-workspace, agentmail, himalaya, scrapling, osint-investigation]\n    homepage: https:\u002F\u002Fgithub.com\u002Fproject-anara\u002Fanara-agent\n---\n\n# unbroker\n\nFind where a person's personal information (name, addresses, phone, email, relatives) is exposed on\ndata brokers and people-search sites, then remove it - automatically where possible, with guided\nhuman steps only where a site demands a CAPTCHA, government ID, phone call, or fax. Manages multiple\npeople independently. It does **not** defeat anti-bot systems, does **not** act on anyone without\nrecorded consent, and does **not** remove public records (voter\u002Fproperty\u002Fcourt) or accounts the\nperson controls.\n\nThe Python CLI (`scripts\u002Fpdd.py`) owns the deterministic state - config, dossiers + consent, the\nbroker database, tier planning, the ledger, drafts, reports, **email sending (SMTP), verification-link\npolling (IMAP), and the autonomous action queue (`next`)**. You (the agent) do the scanning and\nform-driving with native tools: `web_extract` and `browser_navigate` for searching and web forms, and\n`cronjob` for recurring re-scans.\n\n## Autonomy contract\n\nThis skill is designed to run **hands-off**. After intake (+ recorded consent) there are exactly TWO\nlegitimate human touchpoints: (1) the intake conversation itself, and (2) ONE consolidated human-task\ndigest at the end of the run (`$PDD tasks`). Between those:\n\n- **Never ask the operator to choose configuration.** `$PDD setup --auto` detects capabilities and\n  picks the most autonomous valid config itself.\n- **Never pause before individual submissions** when `autonomy=full` (the default): the consent\n  recorded at intake is standing authorization for T0-T2 opt-outs. (`autonomy=assisted` restores\n  per-submission confirmation for cautious operators - honor `confirm_first` flags in `next` output.)\n- **Never interrupt the run for human-only work.** Record it (`record ... human_task_queued\n  --reason \"...\"`) and keep going; it all surfaces once in the final digest.\n- **Drive the whole run as a loop over `$PDD next \u003Csubject>`** - it returns the exact ordered actions\n  to take right now (scan, poll verification, re-check, opt out parents-first, requeue blocked), plus\n  the human digest. Execute every action, record outcomes, re-run `next`, repeat until\n  `done_for_now`. Then present the digest, report, and schedule the cron.\n\nThe hard limits that autonomy never overrides: no acting without recorded consent, no disclosure\nbeyond `disclosure_fields`, no CAPTCHA\u002Fanti-bot bypass, and `confirmed_removed` only after a\nverifying re-scan.\n\n## When to Use\n\n- \"Remove my (or my family member's) data from data brokers \u002F people-search sites.\"\n- \"Opt me out\", \"delete me from Spokeo\u002FWhitepages\u002Fetc.\", \"clean up after a doxxing.\"\n- \"Set up recurring privacy monitoring\" (brokers re-list people).\n- Checking which brokers still expose someone and why.\n\n## Prerequisites\n\n- `python` (stdlib only; no extra packages needed for the core engine).\n- **Optional upgrades** (the skill works zero-config without these; `setup --auto` turns on every\n  one it detects, reading credentials from the shell env **and from `$ANARA_HOME\u002F.env`** so keys\n  Anara already loads for its own tools are picked up without re-exporting - each one converts a\n  class of human tasks into agent actions):\n  - **Cloud browser (recommended default): `BROWSERBASE_API_KEY`.** `setup --auto` selects it\n    whenever the key is present, and it is the intended baseline: a real residential-IP cloud\n    browser **clears soft\u002Fmanaged CAPTCHAs (Cloudflare Turnstile, hCaptcha\u002FreCAPTCHA checkbox) as\n    normal operation**, so those brokers stay automated (T1) instead of becoming human tasks. This\n    is not CAPTCHA \"solving\" - no solver service, no fingerprint spoofing; only interactive\u002Fbehavioral\n    (\"hard\") challenges the browser genuinely cannot pass fall back to a human task. Without the key,\n    the plain agent browser is used and soft-CAPTCHA brokers drop to T2 (human).\n  - Email automation, two credential-free-or-not options:\n    - **Browser mode (no password): `setup --email-mode browser`.** The agent sends opt-out\u002FCCPA\n      emails and opens verification links through the operator's **logged-in webmail** using\n      `browser_*` tools. Nothing is stored. This requires Anara to be pointed at the operator's own\n      logged-in browser, **NOT** a cloud browser: a headless cloud browser (Browserbase) holds no\n      webmail session and is itself Cloudflare\u002FDataDome-gated on webmail and on session-bound broker\n      gates (e.g. PeopleConnect guided-mode). Drive the operator's real Chrome over CDP - launch\n      `chrome --remote-debugging-port=9222 --user-data-dir=\"$HOME\u002F.anara\u002Fchrome-debug\"` (a dedicated\n      debug profile signed into the webmail once, not the Default profile) and connect the browser\n      tools to `127.0.0.1:9222`. **`$PDD cdp` launches this for you** (finds Chrome\u002FChromium\u002FBrave\u002FEdge,\n      starts it detached on the dedicated profile, prints the CDP endpoint; `--check` to test, `--print`\n      for the command). See `references\u002Fmethods.md` -> \"Browser backends: scan vs execute\".\n      Falls back to drafts for an email if the inbox isn't reachable.\n    - **SMTP\u002FIMAP (stored creds): `EMAIL_ADDRESS` + `EMAIL_PASSWORD`** (+ `EMAIL_SMTP_HOST` \u002F\n      `EMAIL_IMAP_HOST` for non-mainstream providers; gmail\u002Foutlook\u002Fyahoo\u002Ficloud\u002Ffastmail inferred).\n      The CLI sends via `send-email` and reads verify links via `poll-verification`. The `agentmail`\n      skill (per-broker aliases) also counts.\n  - Google Sheets tracker: the `google-workspace` skill.\n  - The `scrapling` skill for stealth\u002FCloudflare-protected pages.\n\n## How to Run\n\nRun everything through the `terminal` tool. From this skill's directory:\n\n```bash\nPDD=\"python scripts\u002Fpdd.py\"\n```\n\nThe engine stores data under `$PDD_DATA_DIR` (default `$ANARA_HOME\u002Funbroker`), written\n`0600`. Run via `terminal`, **not** `execute_code` (that sandbox scrubs env and redacts output, which\nbreaks reading the dossier).\n\n## Quick Reference\n\n| Command | Purpose |\n|---|---|\n| `$PDD setup --auto` | **Autonomous setup**: detect capabilities, pick the most autonomous valid config (no questions) |\n| `$PDD doctor` | Readiness check: config, broker count, and which upgrades are on\u002Favailable |\n| `$PDD cdp [--check] [--print] [--port N]` | Launch\u002Fdetect the operator's Chrome over CDP for Phase-2 browser + webmail (dedicated debug profile; the reliable way to send webmail and clear session-bound gates) |\n| `$PDD intake --full-name \"...\" [--alias ...] [--email ... --phone ...] [--city --state] [--prior-location \"City,ST\"] --consent` | Create a consenting subject; captures aliases + multiple emails\u002Fphones + prior locations; prints `subject_id` |\n| `$PDD next \u003Csubject>` | **The autonomous loop driver**: ordered agent actions right now + human digest + `next_wake_at` |\n| `$PDD brokers [--priority crucial]` | List the people-search broker database (curated + live) |\n| `$PDD refresh-brokers` | Pull the latest BADBOOL people-search list **and the CA Data Broker Registry** (`next` requeues this automatically when the cache is stale) |\n| `$PDD registry [--search NAME]` | State registry coverage (CA ~545 ingested; VT\u002FOR\u002FTX portals surfaced); the DROP\u002Femail lane, not scanned |\n| `$PDD drop \u003Csubject> [--filed]` | **The one-shot legal lever**: one CA DROP request deletes from ALL registered brokers; `--filed` records it |\n| `$PDD plan \u003Csubject> [--priority crucial]` | Per-broker tier + method + `search_vectors` + the exact fields to disclose |\n| `$PDD plan \u003Csubject> --batch` | **Reduce view**: overlays ledger state, groups brokers by next action (unscanned\u002Ffound\u002Findirect\u002Fblocked\u002Fin_progress\u002Fdone), collapses ownership clusters, **orders `found` cluster-parents-first + emits a tailored `parent_playbook`**, prints `next_actions` |\n| `$PDD fanout \u003Csubject> [--priority crucial] [--size 5]` | Batch brokers into parallel `delegate_task` subagents (auto for large runs; batches of 5 - 8+ time out) |\n| `$PDD record \u003Csubject> \u003Cbroker> \u003Cstate> [--found true] [--evidence JSON] [--disclosed F --channel C] [--reason \"...\"]` | Update the ledger (validated state machine); **auto-stamps `next_recheck_at`** |\n| `$PDD show \u003Csubject> \u003Cbroker>` | Read back a case's recorded state + evidence + disclosure log (so the parent re-verifies a subagent's `found` without re-deriving the listing URL) |\n| `$PDD send-email \u003Csubject> \u003Cbroker> --listing \u003Curl> [--kind ccpa_indirect ...]` | Render + record the request (recipient locked to the broker's own address). **browser** mode returns a `compose` payload to send via webmail (no password); **programmatic** mode SMTP-sends |\n| `$PDD verify-link \u003Csubject> \u003Cbroker> --text '\u003Cbody>'` | **browser mode**: extract a broker's verification link from webmail text you read (anti-phishing scored) |\n| `$PDD poll-verification \u003Csubject> [--broker \u003Cid>]` | **programmatic mode**: poll IMAP for verification links (anti-phishing scored); auto-advances `submitted → verification_pending` |\n| `$PDD render-email \u003Csubject> \u003Cbroker> --listing \u003Curl>` | Draft only (fallback when no email mode is configured) |\n| `$PDD due \u003Csubject>` | Cases whose recheck window arrived (the cron re-scan queue) |\n| `$PDD tasks \u003Csubject>` | ONE consolidated human-task digest (present at END of run) |\n| `$PDD status \u003Csubject>` | Markdown status report |\n| `$PDD report \u003Csubject> --sheets` | Rows for the Google Sheets tracker |\n\n## Batch operation (two-phase: crawl-all, then delete)\n\nFor anything past a couple of brokers, run this as **map → reduce → act**, not broker-by-broker:\n\n- **Phase 1 - DISCOVER (read-only, parallel, idempotent).** Crawl *every* broker first and record a\n  verdict for each (`found` \u002F `not_found` \u002F `indirect_exposure` \u002F `blocked`). Scanning has no side\n  effects, so it is safe to parallelize and retry. Getting the full exposure map *before* acting is\n  what unlocks cluster dedup and prioritization below. **Default: the parent drives `web_extract`\n  probes directly** - most people-search sites render name\u002Fphone\u002Faddress results as static HTML that\n  `web_extract` reads in seconds. Escalate to `browser_*` only for the few JS-only sites, and to\n  `delegate_task` subagents only for genuinely *reasoning*-heavy work (large-scale namesake\u002Frelative\n  disambiguation). **Do NOT hand a browser-toolset subagent a big list of brokers to crawl** - in the\n  field this timed out repeatedly (600s, ~5-6 brokers each, no summary) because browser navigation is\n  heavy; the ledger writes that survived came at 10x the cost of parent `web_extract`. A `blocked`\n  (DataDome\u002FCloudflare\u002F`antibot`) site is *not* a subagent job either: record `blocked` and requeue it\n  for a stealth\u002Fcloud browser (Browserbase) pass. Subagent reports are self-reports - the parent\n  re-fetches key URLs to confirm a `found` before trusting it (this cuts both ways: it caught a real\n  listing the parent had wrongly assumed was a false positive).\n- **REDUCE - `$PDD plan \u003Csubject> --batch`.** Collapses the crawl into a phase-oriented plan: groups by\n  next action, **collapses ownership clusters** (a parent removal that clears children is ONE action,\n  not N - e.g. one Intelius\u002FPeopleConnect suppression covers Truthfinder\u002FInstant Checkmate\u002FUS Search\u002F…),\n  and prints `next_actions`. `phase` is `discover` while anything is unscanned, else `delete`.\n- **Phase 2 - DELETE (sequential, irreversible).** Work the reduced groups **parents first**:\n  `plan --batch` orders the `found` group cluster-parents-first (most children first) and emits a\n  `parent_playbook` with tailored, ordered steps per parent - follow that order and those steps\n  (full recipes in `references\u002Fmethods.md` → \"Ownership clusters - DO PARENTS FIRST\"). Do the\n  cluster parents (skipping the covered children), **re-scan each parent's children after it confirms**\n  (they usually drop out), then the standalone listings; send the `indirect_exposure` cases as\n  CCPA\u002FGDPR delete-my-PII emails (`send-email --kind ccpa_indirect`), and defer `blocked` to the\n  stealth-browser pass. Opt-outs hit CAPTCHAs, email-verification loops, and session binding - work\n  them **one at a time, carefully** (this is the opposite of fan-out), but do NOT stop to ask\n  permission per submission in `autonomy=full`; in `assisted`, confirm each one. **Usually prefer\n  deletion over suppression** where a broker offers both (Spokeo\u002FBeenVerified) - but follow the\n  record's `deletion.prefer`: **PeopleConnect is the exception** (`prefer: false`), where deleting\n  your user data removes your suppressions and does not stop public-records re-listing, so you\n  suppress-and-maintain instead.\n- **Blind opt-out is the DEFAULT, not a fallback.** Submit an opt-out\u002Fdeletion on **every site with an\n  accessible removal channel, even when a listing was not first confirmed** - it discloses only the\n  subject's own identifiers to the broker's own official channel, so it does not violate\n  least-disclosure. Two corollaries: (1) a guided flow that matches email+DOB+name and says \"no results\"\n  is a **stronger `not_found`** than any scrape - the opt-out flow doubles as the search; (2) when a form\n  is automation-hostile (hard CAPTCHA, Cloudflare\u002FDataDome, slide-to-verify slider), **default to the\n  broker's cited rights-request email** (name+state+contact-email only) rather than recording `blocked`.\n  CAPTCHA policy: never defeat behavioral\u002Ftoken\u002Fslider challenges; OK to read a static distorted-text or\n  plain-arithmetic CAPTCHA on the subject's own opt-out, but stop if the site rejects the whole\n  submission after a correct answer (it is fingerprinting the automation). Third-party\u002Findirect records\n  are the exception - still confirm those before acting. Per-site game plans + the meta-search no-op\n  skip-list are in `references\u002Fsite-playbooks.md`; the full policy is in `references\u002Fmethods.md`.\n- **PeopleConnect delete-wipes-suppression (permanent rule).** A PeopleConnect *deletion* wipes the\n  suppression and the subject re-lists across the whole affiliate cluster. If a \"Your deletion request\n  for PeopleConnect.us is Complete\" email ever appears, the suppression is gone -> **re-run suppression\n  and re-verify** the Control step reads \"suppressed\". Never leave this cluster on a completed deletion\n  (see `references\u002Fbrokers\u002Fintelius.json`).\n\nSubagent reports are self-reports: the parent re-verifies key claims (listing URLs, match basis) before\nrecording `found` and before any deletion.\n\n## Procedure (the autonomous loop)\n\n1. **Setup (once, no questions).** Run `$PDD setup --auto` - it detects capabilities and configures\n   the most autonomous valid combination itself (programmatic email when `EMAIL_*` creds exist,\n   Browserbase when its key exists, `age` encryption when the binary exists, `autonomy=full`). Then\n   `$PDD doctor` and show the operator the readiness output **for information, not as a question** -\n   proceed immediately. Mention what would unlock more automation (e.g. email creds) but do not wait.\n2. **Intake + consent (the ONE human conversation).** `$PDD intake ...` with `--consent` (and\n   `--consent-method`). Without consent the engine refuses to plan or act. Collect everything in one\n   pass - names\u002Faliases, current + prior cities, emails, phones - so you never have to come back with\n   questions. For California subjects, also read `references\u002Flegal\u002Fdrop.md`: `next` will surface a\n   `drop_submit` one-shot that deletes from every registered broker (~545) at once, which is the\n   single highest-leverage action. File it, then `drop \u003Csubject> --filed`. For non-CA subjects the\n   registry is covered by targeted CCPA\u002FGDPR emails (`registry --search`, then `send-email`); the\n   people-search sites are worked directly in either case.\n3. **Drain the queue.** Loop:\n\n   ```\n   while true:\n     q = $PDD next \u003Csubject>\n     if q.actions is empty: break\n     execute EVERY action in order; record each outcome via $PDD record\n   ```\n\n   `next` emits, in order: `refresh_brokers` (stale cache), `fanout_scan`\u002F`scan_inline` (Phase 1\n   crawl - see step 4), `poll_verification` (in-flight email confirmations), `verify_removal` (due\n   re-checks), `optout_web_form`\u002F`optout_email_send` (Phase 2, parents-first with playbook steps),\n   `indirect_email_send`, and `stealth_rescan`. Human-only work never appears as an action - it\n   accumulates in `q.human_digest`. In `autonomy=full`, execute actions without pausing; honor\n   `confirm_first` in `assisted` mode.\n4. **Scanning (when `next` says so).** For `fanout_scan`: run `$PDD fanout \u003Csubject>` and **spawn one\n   `delegate_task` subagent per `batch`, in parallel, passing that batch's ready-made `brief`** - do\n   not scan all brokers yourself sequentially. For `scan_inline`: scan the few brokers yourself.\n   Either way, each broker gets **every** `search_vectors` entry via the `references\u002Fmethods.md`\n   ladder (`web_extract` → `site:` probe → `browser_navigate` → `scrapling`), a 404 is INCONCLUSIVE\n   (not `not_found`), `blocked` is recorded when `antibot` is set and no stealth browser is available,\n   and subject vs namesake\u002Frelative is confirmed before recording:\n   `$PDD record \u003Csubject> \u003Cbroker> \u003Cfound|not_found|indirect_exposure|blocked> --found \u003Cbool> --evidence '{\"listing_urls\":[...]}'`.\n   The parent re-verifies key `found` claims from subagents before trusting them.\n5. **Opt-outs (when `next` says so).** Actions come pre-ordered parents-first with `steps` from each\n   broker record's own `optout.playbook` (field-verified; cluster parents like PeopleConnect,\n   Whitepages, BeenVerified, Spokeo have exact, live-checked recipes). **Deletion usually beats\n   suppression**: when an action carries `prefer_deletion`, complete the record's DELETION lane, not\n   just the hide-my-listing flow. When it carries `prefer_suppression` instead (**PeopleConnect** -\n   deleting removes your suppressions and does not stop re-listing), do the suppression flow and keep\n   it maintained; use their Delete button only for a deliberate data-purge. Per method:\n   - **web_form** → drive `optout_url` with `browser_navigate`\u002F`browser_type`\u002F`browser_click`, submit\n     only `disclosure_fields`, screenshot the confirmation, then the action's `after` record command.\n     Playbooks may end with a right-to-delete `send-email` follow-up - do it (full erasure, not just\n     listing suppression).\n    - **email** → `$PDD send-email \u003Csubject> \u003Cbroker> --kind \u003Cccpa|gdpr|generic> --to \u003Caddr>\n      --listing \u003Curl>` records + discloses in one step (recipient locked to addresses the broker\n      record declares; `next` picks the kind from residency - never claim CCPA\u002FGDPR for someone who\n      can't). In **browser** mode it returns a recipient-locked `compose` payload: compose a new\n      message to `compose.to` with `compose.subject`\u002F`compose.body` exactly in the operator's webmail\n      via `browser_*` and send (no password); in **programmatic** mode it SMTP-sends. `next` also\n      routes human-gated forms (phone-callback\u002Fgov-ID) through a broker's deletion email when one\n      exists - the **rescue lane** (verified Whitepages pattern). Draft-only falls back to\n      `render-email` + a digest entry.\n   - **captcha** → soft\u002Fmanaged challenges clear automatically on the default cloud browser (proceed\n     as normal); only a hard interactive\u002Fbehavioral challenge it can't pass is recorded `blocked`\n     (requeued for the stealth\u002Foperator-browser pass). Never a solver service.\n   - **phone_callback \u002F account \u002F gov_id \u002F fax \u002F mail \u002F voice (T3)** *without a deletion email* →\n     never an agent action; `next` already routed these to the digest. Record them:\n     `$PDD record \u003Csubject> \u003Cbroker> human_task_queued --reason \"...\"`.\n 6. **Verification (when `next` says so).** In **programmatic** mode `$PDD poll-verification \u003Csubject>`\n    finds arrived confirmation links via IMAP (anti-phishing scored, auto-advances state). In\n    **browser** mode, open the broker's confirmation email in the operator's webmail and run\n    `$PDD verify-link \u003Csubject> \u003Cbroker> --text '\u003Cbody>'` to score the link. Either way **open the\n    link in the same browser** (several brokers bind the verification session to the browser that\n    opens it), finish the flow, then record `awaiting_processing`. `confirmed_removed` ONLY after a\n    verifying re-scan shows the listing gone - never off the submission flow's own confirmation page.\n7. **Wrap up (once per run).** When `next` returns no actions: present `$PDD tasks \u003Csubject>` (the\n   consolidated human digest) if non-empty, then `$PDD status \u003Csubject>`; if the Sheets tracker is\n   on, append `$PDD report \u003Csubject> --sheets` rows via the `google-workspace` skill.\n8. **Schedule the next wake-up.** `next` returns `next_wake_at` (earliest due re-check). Create ONE\n   `cronjob` that re-runs this skill's loop for the subject (a prompt like: *\"run the\n   unbroker loop for \u003Csubject_id>: `$PDD next` and execute all actions\"*). Processing\n   windows, verification polls, and reappearance sweeps all flow through the same queue, so the case\n   keeps advancing with zero human attention.\n\n## Pitfalls\n\n- **Never disclose more than the broker already shows.** Submit only `disclosure_fields`. The engine\n  never volunteers SSN\u002FID numbers; you must not either.\n- **No consent, no action.** The engine enforces this; do not work around it to \"research\" a third party.\n- **`send-email` is idempotent + rate-limited.** It refuses to re-send a case already `submitted`\n  or beyond (use `--force` only if a genuine re-send is needed), and SMTP sends are paced by\n  `email_min_interval_seconds` (default 20s) with retry\u002Fbackoff. Do not loop it to \"make sure\" -\n  a successful SMTP handoff is not proof of delivery; the due-queue re-scan is the real confirmation.\n- **Ledger writes are locked.** Concurrent runs (cron + manual) serialize safely; if you ever see a\n  lock timeout, another run is mid-write - let it finish, don't delete the `.lock` by hand.\n- **Autonomy ≠ improvisation.** Full autonomy means not *asking* between steps; it does not loosen any\n  gate. If a broker demands MORE than the planned `disclosure_fields` mid-flow, stop that case and\n  queue it (`human_task_queued --reason`) rather than deciding alone to disclose extra PII.\n- **Don't interrupt the run with questions.** Config choices are `setup --auto`'s job; human-only work\n  goes to the digest. The only mid-run question that's ever warranted is a missing-identity fact that\n  blocks scanning (e.g. no city at all) - and that should have been collected at intake.\n- **Use `terminal`, not `execute_code`** for `pdd.py` (secret scrubbing + output redaction break it).\n- **Dossiers are plaintext by default** (JSON, `0600` under `ANARA_HOME`). For at-rest encryption run\n  `$PDD setup --encryption age` - it generates a local `age` key and encrypts dossiers + ledgers (the\n  audit log holds field names only and stays plaintext). It guards casual\u002Fbackup\u002Fcommit exposure, not\n  a full-`ANARA_HOME` read; set `PDD_AGE_IDENTITY` to a separate volume for real key separation.\n  `$PDD doctor` shows whether encryption is *actually* engaged (not just whether `age` is installed).\n- **\"Hidden from free search\" ≠ deleted.** Only mark `confirmed_removed` after verifying the record is\n  actually gone; note paid-tier retention in the report.\n- **Soft CAPTCHAs clear by default; don't fight the hard ones.** The default cloud browser passes\n  managed\u002Fsoft challenges as normal operation (those brokers stay T1). For a hard interactive one it\n  genuinely can't pass, record `blocked` and let the stealth\u002Foperator-browser pass take it - never a\n  third-party solver service or fingerprint spoofing.\n- **Broker pages change.** If a flow breaks, `$PDD record ... blocked` and flag the broker file in\n  `references\u002Fbrokers\u002F` for re-verification instead of guessing.\n- **Verify non-field-verified records before submitting.** `confidence: auto` records came from\n  parsing BADBOOL (read `optout.notes`\u002F`optout.links`, confirm the real opt-out URL). `confidence:\n  documented` records (several people-search sites) carry the correct published opt-out URL but have\n  **not** been field-verified (they 403 datacenter IPs), so confirm the live flow via the operator's\n  residential browser on first use, then set `last_verified`. Field-verified curated records (no\n  `confidence`, e.g. the cluster parents) have checked mechanics and take precedence.\n\n## Verification\n\n- `scripts\u002Frun_tests.sh tests\u002Fskills\u002Ftest_unbroker_skill.py` (hermetic; no network), or the\n  dependency-free runner `python tests\u002Fskills\u002Ftest_unbroker_skill.py`.\n- Dry run: `$PDD setup --auto && $PDD doctor && SID=$($PDD intake --full-name \"Test Person\"\n  --email t@example.com --consent | python -c 'import sys,json;print(json.load(sys.stdin)[\"subject_id\"])')\n  && $PDD next \"$SID\"` and confirm a readiness summary plus an ordered action queue.\n","# unbroker\n\nFind where a person's personal information (name, addresses, phone, email, relatives) is exposed on\ndata brokers and people-search sites, then remove it - automatically where possible, with guided\nhuman steps only where a site demands a CAPTCHA, government ID, phone call, or fax. Manages multiple\npeople independently. It does **not** defeat anti-bot systems, does **not** act on anyone without\nrecorded consent, and does **not** remove public records (voter\u002Fproperty\u002Fcourt) or accounts the\nperson controls.\n\nThe Python CLI (`scripts\u002Fpdd.py`) owns the deterministic state - config, dossiers + consent, the\nbroker database, tier planning, the ledger, drafts, reports, **email sending (SMTP), verification-link\npolling (IMAP), and the autonomous action queue (`next`)**. You (the agent) do the scanning and\nform-driving with native tools: `web_extract` and `browser_navigate` for searching and web forms, and\n`cronjob` for recurring re-scans.\n\n## Autonomy contract\n\nThis skill is designed to run **hands-off**. After intake (+ recorded consent) there are exactly TWO\nlegitimate human touchpoints: (1) the intake conversation itself, and (2) ONE consolidated human-task\ndigest at the end of the run (`$PDD tasks`). Between those:\n\n- **Never ask the operator to choose configuration.** `$PDD setup --auto` detects capabilities and\n  picks the most autonomous valid config itself.\n- **Never pause before individual submissions** when `autonomy=full` (the default): the consent\n  recorded at intake is standing authorization for T0-T2 opt-outs. (`autonomy=assisted` restores\n  per-submission confirmation for cautious operators - honor `confirm_first` flags in `next` output.)\n- **Never interrupt the run for human-only work.** Record it (`record ... human_task_queued\n  --reason \"...\"`) and keep going; it all surfaces once in the final digest.\n- **Drive the whole run as a loop over `$PDD next \u003Csubject>`** - it returns the exact ordered actions\n  to take right now (scan, poll verification, re-check, opt out parents-first, requeue blocked), plus\n  the human digest. Execute every action, record outcomes, re-run `next`, repeat until\n  `done_for_now`. Then present the digest, report, and schedule the cron.\n\nThe hard limits that autonomy never overrides: no acting without recorded consent, no disclosure\nbeyond `disclosure_fields`, no CAPTCHA\u002Fanti-bot bypass, and `confirmed_removed` only after a\nverifying re-scan.\n\n## When to Use\n\n- \"Remove my (or my family member's) data from data brokers \u002F people-search sites.\"\n- \"Opt me out\", \"delete me from Spokeo\u002FWhitepages\u002Fetc.\", \"clean up after a doxxing.\"\n- \"Set up recurring privacy monitoring\" (brokers re-list people).\n- Checking which brokers still expose someone and why.\n\n## Prerequisites\n\n- `python` (stdlib only; no extra packages needed for the core engine).\n- **Optional upgrades** (the skill works zero-config without these; `setup --auto` turns on every\n  one it detects, reading credentials from the shell env **and from `$ANARA_HOME\u002F.env`** so keys\n  Anara already loads for its own tools are picked up without re-exporting - each one converts a\n  class of human tasks into agent actions):\n  - **Cloud browser (recommended default): `BROWSERBASE_API_KEY`.** `setup --auto` selects it\n    whenever the key is present, and it is the intended baseline: a real residential-IP cloud\n    browser **clears soft\u002Fmanaged CAPTCHAs (Cloudflare Turnstile, hCaptcha\u002FreCAPTCHA checkbox) as\n    normal operation**, so those brokers stay automated (T1) instead of becoming human tasks. This\n    is not CAPTCHA \"solving\" - no solver service, no fingerprint spoofing; only interactive\u002Fbehavioral\n    (\"hard\") challenges the browser genuinely cannot pass fall back to a human task. Without the key,\n    the plain agent browser is used and soft-CAPTCHA brokers drop to T2 (human).\n  - Email automation, two credential-free-or-not options:\n    - **Browser mode (no password): `setup --email-mode browser`.** The agent sends opt-out\u002FCCPA\n      emails and opens verification links through the operator's **logged-in webmail** using\n      `browser_*` tools. Nothing is stored. This requires Anara to be pointed at the operator's own\n      logged-in browser, **NOT** a cloud browser: a headless cloud browser (Browserbase) holds no\n      webmail session and is itself Cloudflare\u002FDataDome-gated on webmail and on session-bound broker\n      gates (e.g. PeopleConnect guided-mode). Drive the operator's real Chrome over CDP - launch\n      `chrome --remote-debugging-port=9222 --user-data-dir=\"$HOME\u002F.anara\u002Fchrome-debug\"` (a dedicated\n      debug profile signed into the webmail once, not the Default profile) and connect the browser\n      tools to `127.0.0.1:9222`. **`$PDD cdp` launches this for you** (finds Chrome\u002FChromium\u002FBrave\u002FEdge,\n      starts it detached on the dedicated profile, prints the CDP endpoint; `--check` to test, `--print`\n      for the command). See `references\u002Fmethods.md` -> \"Browser backends: scan vs execute\".\n      Falls back to drafts for an email if the inbox isn't reachable.\n    - **SMTP\u002FIMAP (stored creds): `EMAIL_ADDRESS` + `EMAIL_PASSWORD`** (+ `EMAIL_SMTP_HOST` \u002F\n      `EMAIL_IMAP_HOST` for non-mainstream providers; gmail\u002Foutlook\u002Fyahoo\u002Ficloud\u002Ffastmail inferred).\n      The CLI sends via `send-email` and reads verify links via `poll-verification`. The `agentmail`\n      skill (per-broker aliases) also counts.\n  - Google Sheets tracker: the `google-workspace` skill.\n  - The `scrapling` skill for stealth\u002FCloudflare-protected pages.\n\n## How to Run\n\nRun everything through the `terminal` tool. From this skill's directory:\n\n```bash\nPDD=\"python scripts\u002Fpdd.py\"\n```\n\nThe engine stores data under `$PDD_DATA_DIR` (default `$ANARA_HOME\u002Funbroker`), written\n`0600`. Run via `terminal`, **not** `execute_code` (that sandbox scrubs env and redacts output, which\nbreaks reading the dossier).\n\n## Quick Reference\n\n| Command | Purpose |\n|---|---|\n| `$PDD setup --auto` | **Autonomous setup**: detect capabilities, pick the most autonomous valid config (no questions) |\n| `$PDD doctor` | Readiness check: config, broker count, and which upgrades are on\u002Favailable |\n| `$PDD cdp [--check] [--print] [--port N]` | Launch\u002Fdetect the operator's Chrome over CDP for Phase-2 browser + webmail (dedicated debug profile; the reliable way to send webmail and clear session-bound gates) |\n| `$PDD intake --full-name \"...\" [--alias ...] [--email ... --phone ...] [--city --state] [--prior-location \"City,ST\"] --consent` | Create a consenting subject; captures aliases + multiple emails\u002Fphones + prior locations; prints `subject_id` |\n| `$PDD next \u003Csubject>` | **The autonomous loop driver**: ordered agent actions right now + human digest + `next_wake_at` |\n| `$PDD brokers [--priority crucial]` | List the people-search broker database (curated + live) |\n| `$PDD refresh-brokers` | Pull the latest BADBOOL people-search list **and the CA Data Broker Registry** (`next` requeues this automatically when the cache is stale) |\n| `$PDD registry [--search NAME]` | State registry coverage (CA ~545 ingested; VT\u002FOR\u002FTX portals surfaced); the DROP\u002Femail lane, not scanned |\n| `$PDD drop \u003Csubject> [--filed]` | **The one-shot legal lever**: one CA DROP request deletes from ALL registered brokers; `--filed` records it |\n| `$PDD plan \u003Csubject> [--priority crucial]` | Per-broker tier + method + `search_vectors` + the exact fields to disclose |\n| `$PDD plan \u003Csubject> --batch` | **Reduce view**: overlays ledger state, groups brokers by next action (unscanned\u002Ffound\u002Findirect\u002Fblocked\u002Fin_progress\u002Fdone), collapses ownership clusters, **orders `found` cluster-parents-first + emits a tailored `parent_playbook`**, prints `next_actions` |\n| `$PDD fanout \u003Csubject> [--priority crucial] [--size 5]` | Batch brokers into parallel `delegate_task` subagents (auto for large runs; batches of 5 - 8+ time out) |\n| `$PDD record \u003Csubject> \u003Cbroker> \u003Cstate> [--found true] [--evidence JSON] [--disclosed F --channel C] [--reason \"...\"]` | Update the ledger (validated state machine); **auto-stamps `next_recheck_at`** |\n| `$PDD show \u003Csubject> \u003Cbroker>` | Read back a case's recorded state + evidence + disclosure log (so the parent re-verifies a subagent's `found` without re-deriving the listing URL) |\n| `$PDD send-email \u003Csubject> \u003Cbroker> --listing \u003Curl> [--kind ccpa_indirect ...]` | Render + record the request (recipient locked to the broker's own address). **browser** mode returns a `compose` payload to send via webmail (no password); **programmatic** mode SMTP-sends |\n| `$PDD verify-link \u003Csubject> \u003Cbroker> --text '\u003Cbody>'` | **browser mode**: extract a broker's verification link from webmail text you read (anti-phishing scored) |\n| `$PDD poll-verification \u003Csubject> [--broker \u003Cid>]` | **programmatic mode**: poll IMAP for verification links (anti-phishing scored); auto-advances `submitted → verification_pending` |\n| `$PDD render-email \u003Csubject> \u003Cbroker> --listing \u003Curl>` | Draft only (fallback when no email mode is configured) |\n| `$PDD due \u003Csubject>` | Cases whose recheck window arrived (the cron re-scan queue) |\n| `$PDD tasks \u003Csubject>` | ONE consolidated human-task digest (present at END of run) |\n| `$PDD status \u003Csubject>` | Markdown status report |\n| `$PDD report \u003Csubject> --sheets` | Rows for the Google Sheets tracker |\n\n## Batch operation (two-phase: crawl-all, then delete)\n\nFor anything past a couple of brokers, run this as **map → reduce → act**, not broker-by-broker:\n\n- **Phase 1 - DISCOVER (read-only, parallel, idempotent).** Crawl *every* broker first and record a\n  verdict for each (`found` \u002F `not_found` \u002F `indirect_exposure` \u002F `blocked`). Scanning has no side\n  effects, so it is safe to parallelize and retry. Getting the full exposure map *before* acting is\n  what unlocks cluster dedup and prioritization below. **Default: the parent drives `web_extract`\n  probes directly** - most people-search sites render name\u002Fphone\u002Faddress results as static HTML that\n  `web_extract` reads in seconds. Escalate to `browser_*` only for the few JS-only sites, and to\n  `delegate_task` subagents only for genuinely *reasoning*-heavy work (large-scale namesake\u002Frelative\n  disambiguation). **Do NOT hand a browser-toolset subagent a big list of brokers to crawl** - in the\n  field this timed out repeatedly (600s, ~5-6 brokers each, no summary) because browser navigation is\n  heavy; the ledger writes that survived came at 10x the cost of parent `web_extract`. A `blocked`\n  (DataDome\u002FCloudflare\u002F`antibot`) site is *not* a subagent job either: record `blocked` and requeue it\n  for a stealth\u002Fcloud browser (Browserbase) pass. Subagent reports are self-reports - the parent\n  re-fetches key URLs to confirm a `found` before trusting it (this cuts both ways: it caught a real\n  listing the parent had wrongly assumed was a false positive).\n- **REDUCE - `$PDD plan \u003Csubject> --batch`.** Collapses the crawl into a phase-oriented plan: groups by\n  next action, **collapses ownership clusters** (a parent removal that clears children is ONE action,\n  not N - e.g. one Intelius\u002FPeopleConnect suppression covers Truthfinder\u002FInstant Checkmate\u002FUS Search\u002F…),\n  and prints `next_actions`. `phase` is `discover` while anything is unscanned, else `delete`.\n- **Phase 2 - DELETE (sequential, irreversible).** Work the reduced groups **parents first**:\n  `plan --batch` orders the `found` group cluster-parents-first (most children first) and emits a\n  `parent_playbook` with tailored, ordered steps per parent - follow that order and those steps\n  (full recipes in `references\u002Fmethods.md` → \"Ownership clusters - DO PARENTS FIRST\"). Do the\n  cluster parents (skipping the covered children), **re-scan each parent's children after it confirms**\n  (they usually drop out), then the standalone listings; send the `indirect_exposure` cases as\n  CCPA\u002FGDPR delete-my-PII emails (`send-email --kind ccpa_indirect`), and defer `blocked` to the\n  stealth-browser pass. Opt-outs hit CAPTCHAs, email-verification loops, and session binding - work\n  them **one at a time, carefully** (this is the opposite of fan-out), but do NOT stop to ask\n  permission per submission in `autonomy=full`; in `assisted`, confirm each one. **Usually prefer\n  deletion over suppression** where a broker offers both (Spokeo\u002FBeenVerified) - but follow the\n  record's `deletion.prefer`: **PeopleConnect is the exception** (`prefer: false`), where deleting\n  your user data removes your suppressions and does not stop public-records re-listing, so you\n  suppress-and-maintain instead.\n- **Blind opt-out is the DEFAULT, not a fallback.** Submit an opt-out\u002Fdeletion on **every site with an\n  accessible removal channel, even when a listing was not first confirmed** - it discloses only the\n  subject's own identifiers to the broker's own official channel, so it does not violate\n  least-disclosure. Two corollaries: (1) a guided flow that matches email+DOB+name and says \"no results\"\n  is a **stronger `not_found`** than any scrape - the opt-out flow doubles as the search; (2) when a form\n  is automation-hostile (hard CAPTCHA, Cloudflare\u002FDataDome, slide-to-verify slider), **default to the\n  broker's cited rights-request email** (name+state+contact-email only) rather than recording `blocked`.\n  CAPTCHA policy: never defeat behavioral\u002Ftoken\u002Fslider challenges; OK to read a static distorted-text or\n  plain-arithmetic CAPTCHA on the subject's own opt-out, but stop if the site rejects the whole\n  submission after a correct answer (it is fingerprinting the automation). Third-party\u002Findirect records\n  are the exception - still confirm those before acting. Per-site game plans + the meta-search no-op\n  skip-list are in `references\u002Fsite-playbooks.md`; the full policy is in `references\u002Fmethods.md`.\n- **PeopleConnect delete-wipes-suppression (permanent rule).** A PeopleConnect *deletion* wipes the\n  suppression and the subject re-lists across the whole affiliate cluster. If a \"Your deletion request\n  for PeopleConnect.us is Complete\" email ever appears, the suppression is gone -> **re-run suppression\n  and re-verify** the Control step reads \"suppressed\". Never leave this cluster on a completed deletion\n  (see `references\u002Fbrokers\u002Fintelius.json`).\n\nSubagent reports are self-reports: the parent re-verifies key claims (listing URLs, match basis) before\nrecording `found` and before any deletion.\n\n## Procedure (the autonomous loop)\n\n1. **Setup (once, no questions).** Run `$PDD setup --auto` - it detects capabilities and configures\n   the most autonomous valid combination itself (programmatic email when `EMAIL_*` creds exist,\n   Browserbase when its key exists, `age` encryption when the binary exists, `autonomy=full`). Then\n   `$PDD doctor` and show the operator the readiness output **for information, not as a question** -\n   proceed immediately. Mention what would unlock more automation (e.g. email creds) but do not wait.\n2. **Intake + consent (the ONE human conversation).** `$PDD intake ...` with `--consent` (and\n   `--consent-method`). Without consent the engine refuses to plan or act. Collect everything in one\n   pass - names\u002Faliases, current + prior cities, emails, phones - so you never have to come back with\n   questions. For California subjects, also read `references\u002Flegal\u002Fdrop.md`: `next` will surface a\n   `drop_submit` one-shot that deletes from every registered broker (~545) at once, which is the\n   single highest-leverage action. File it, then `drop \u003Csubject> --filed`. For non-CA subjects the\n   registry is covered by targeted CCPA\u002FGDPR emails (`registry --search`, then `send-email`); the\n   people-search sites are worked directly in either case.\n3. **Drain the queue.** Loop:\n\n   ```\n   while true:\n     q = $PDD next \u003Csubject>\n     if q.actions is empty: break\n     execute EVERY action in order; record each outcome via $PDD record\n   ```\n\n   `next` emits, in order: `refresh_brokers` (stale cache), `fanout_scan`\u002F`scan_inline` (Phase 1\n   crawl - see step 4), `poll_verification` (in-flight email confirmations), `verify_removal` (due\n   re-checks), `optout_web_form`\u002F`optout_email_send` (Phase 2, parents-first with playbook steps),\n   `indirect_email_send`, and `stealth_rescan`. Human-only work never appears as an action - it\n   accumulates in `q.human_digest`. In `autonomy=full`, execute actions without pausing; honor\n   `confirm_first` in `assisted` mode.\n4. **Scanning (when `next` says so).** For `fanout_scan`: run `$PDD fanout \u003Csubject>` and **spawn one\n   `delegate_task` subagent per `batch`, in parallel, passing that batch's ready-made `brief`** - do\n   not scan all brokers yourself sequentially. For `scan_inline`: scan the few brokers yourself.\n   Either way, each broker gets **every** `search_vectors` entry via the `references\u002Fmethods.md`\n   ladder (`web_extract` → `site:` probe → `browser_navigate` → `scrapling`), a 404 is INCONCLUSIVE\n   (not `not_found`), `blocked` is recorded when `antibot` is set and no stealth browser is available,\n   and subject vs namesake\u002Frelative is confirmed before recording:\n   `$PDD record \u003Csubject> \u003Cbroker> \u003Cfound|not_found|indirect_exposure|blocked> --found \u003Cbool> --evidence '{\"listing_urls\":[...]}'`.\n   The parent re-verifies key `found` claims from subagents before trusting them.\n5. **Opt-outs (when `next` says so).** Actions come pre-ordered parents-first with `steps` from each\n   broker record's own `optout.playbook` (field-verified; cluster parents like PeopleConnect,\n   Whitepages, BeenVerified, Spokeo have exact, live-checked recipes). **Deletion usually beats\n   suppression**: when an action carries `prefer_deletion`, complete the record's DELETION lane, not\n   just the hide-my-listing flow. When it carries `prefer_suppression` instead (**PeopleConnect** -\n   deleting removes your suppressions and does not stop re-listing), do the suppression flow and keep\n   it maintained; use their Delete button only for a deliberate data-purge. Per method:\n   - **web_form** → drive `optout_url` with `browser_navigate`\u002F`browser_type`\u002F`browser_click`, submit\n     only `disclosure_fields`, screenshot the confirmation, then the action's `after` record command.\n     Playbooks may end with a right-to-delete `send-email` follow-up - do it (full erasure, not just\n     listing suppression).\n    - **email** → `$PDD send-email \u003Csubject> \u003Cbroker> --kind \u003Cccpa|gdpr|generic> --to \u003Caddr>\n      --listing \u003Curl>` records + discloses in one step (recipient locked to addresses the broker\n      record declares; `next` picks the kind from residency - never claim CCPA\u002FGDPR for someone who\n      can't). In **browser** mode it returns a recipient-locked `compose` payload: compose a new\n      message to `compose.to` with `compose.subject`\u002F`compose.body` exactly in the operator's webmail\n      via `browser_*` and send (no password); in **programmatic** mode it SMTP-sends. `next` also\n      routes human-gated forms (phone-callback\u002Fgov-ID) through a broker's deletion email when one\n      exists - the **rescue lane** (verified Whitepages pattern). Draft-only falls back to\n      `render-email` + a digest entry.\n   - **captcha** → soft\u002Fmanaged challenges clear automatically on the default cloud browser (proceed\n     as normal); only a hard interactive\u002Fbehavioral challenge it can't pass is recorded `blocked`\n     (requeued for the stealth\u002Foperator-browser pass). Never a solver service.\n   - **phone_callback \u002F account \u002F gov_id \u002F fax \u002F mail \u002F voice (T3)** *without a deletion email* →\n     never an agent action; `next` already routed these to the digest. Record them:\n     `$PDD record \u003Csubject> \u003Cbroker> human_task_queued --reason \"...\"`.\n 6. **Verification (when `next` says so).** In **programmatic** mode `$PDD poll-verification \u003Csubject>`\n    finds arrived confirmation links via IMAP (anti-phishing scored, auto-advances state). In\n    **browser** mode, open the broker's confirmation email in the operator's webmail and run\n    `$PDD verify-link \u003Csubject> \u003Cbroker> --text '\u003Cbody>'` to score the link. Either way **open the\n    link in the same browser** (several brokers bind the verification session to the browser that\n    opens it), finish the flow, then record `awaiting_processing`. `confirmed_removed` ONLY after a\n    verifying re-scan shows the listing gone - never off the submission flow's own confirmation page.\n7. **Wrap up (once per run).** When `next` returns no actions: present `$PDD tasks \u003Csubject>` (the\n   consolidated human digest) if non-empty, then `$PDD status \u003Csubject>`; if the Sheets tracker is\n   on, append `$PDD report \u003Csubject> --sheets` rows via the `google-workspace` skill.\n8. **Schedule the next wake-up.** `next` returns `next_wake_at` (earliest due re-check). Create ONE\n   `cronjob` that re-runs this skill's loop for the subject (a prompt like: *\"run the\n   unbroker loop for \u003Csubject_id>: `$PDD next` and execute all actions\"*). Processing\n   windows, verification polls, and reappearance sweeps all flow through the same queue, so the case\n   keeps advancing with zero human attention.\n\n## Pitfalls\n\n- **Never disclose more than the broker already shows.** Submit only `disclosure_fields`. The engine\n  never volunteers SSN\u002FID numbers; you must not either.\n- **No consent, no action.** The engine enforces this; do not work around it to \"research\" a third party.\n- **`send-email` is idempotent + rate-limited.** It refuses to re-send a case already `submitted`\n  or beyond (use `--force` only if a genuine re-send is needed), and SMTP sends are paced by\n  `email_min_interval_seconds` (default 20s) with retry\u002Fbackoff. Do not loop it to \"make sure\" -\n  a successful SMTP handoff is not proof of delivery; the due-queue re-scan is the real confirmation.\n- **Ledger writes are locked.** Concurrent runs (cron + manual) serialize safely; if you ever see a\n  lock timeout, another run is mid-write - let it finish, don't delete the `.lock` by hand.\n- **Autonomy ≠ improvisation.** Full autonomy means not *asking* between steps; it does not loosen any\n  gate. If a broker demands MORE than the planned `disclosure_fields` mid-flow, stop that case and\n  queue it (`human_task_queued --reason`) rather than deciding alone to disclose extra PII.\n- **Don't interrupt the run with questions.** Config choices are `setup --auto`'s job; human-only work\n  goes to the digest. The only mid-run question that's ever warranted is a missing-identity fact that\n  blocks scanning (e.g. no city at all) - and that should have been collected at intake.\n- **Use `terminal`, not `execute_code`** for `pdd.py` (secret scrubbing + output redaction break it).\n- **Dossiers are plaintext by default** (JSON, `0600` under `ANARA_HOME`). For at-rest encryption run\n  `$PDD setup --encryption age` - it generates a local `age` key and encrypts dossiers + ledgers (the\n  audit log holds field names only and stays plaintext). It guards casual\u002Fbackup\u002Fcommit exposure, not\n  a full-`ANARA_HOME` read; set `PDD_AGE_IDENTITY` to a separate volume for real key separation.\n  `$PDD doctor` shows whether encryption is *actually* engaged (not just whether `age` is installed).\n- **\"Hidden from free search\" ≠ deleted.** Only mark `confirmed_removed` after verifying the record is\n  actually gone; note paid-tier retention in the report.\n- **Soft CAPTCHAs clear by default; don't fight the hard ones.** The default cloud browser passes\n  managed\u002Fsoft challenges as normal operation (those brokers stay T1). For a hard interactive one it\n  genuinely can't pass, record `blocked` and let the stealth\u002Foperator-browser pass take it - never a\n  third-party solver service or fingerprint spoofing.\n- **Broker pages change.** If a flow breaks, `$PDD record ... blocked` and flag the broker file in\n  `references\u002Fbrokers\u002F` for re-verification instead of guessing.\n- **Verify non-field-verified records before submitting.** `confidence: auto` records came from\n  parsing BADBOOL (read `optout.notes`\u002F`optout.links`, confirm the real opt-out URL). `confidence:\n  documented` records (several people-search sites) carry the correct published opt-out URL but have\n  **not** been field-verified (they 403 datacenter IPs), so confirm the live flow via the operator's\n  residential browser on first use, then set `last_verified`. Field-verified curated records (no\n  `confidence`, e.g. the cluster parents) have checked mechanics and take precedence.\n\n## Verification\n\n- `scripts\u002Frun_tests.sh tests\u002Fskills\u002Ftest_unbroker_skill.py` (hermetic; no network), or the\n  dependency-free runner `python tests\u002Fskills\u002Ftest_unbroker_skill.py`.\n- Dry run: `$PDD setup --auto && $PDD doctor && SID=$($PDD intake --full-name \"Test Person\"\n  --email t@example.com --consent | python -c 'import sys,json;print(json.load(sys.stdin)[\"subject_id\"])')\n  && $PDD next \"$SID\"` and confirm a readiness summary plus an ordered action queue.\n",{"name":5,"author":23,"license":14,"version":24,"metadata":25,"platforms":42,"description":7,"prerequisites":46},"SHL0MS (github.com\u002FSHL0MS)","1.0.0",{"anara":26},{"tags":27,"category":33,"homepage":35,"related_skills":36},[28,29,30,31,32,33,34],"privacy","data-broker","opt-out","ccpa","gdpr","security","doxxing","https:\u002F\u002Fgithub.com\u002Fproject-anara\u002Fanara-agent",[37,38,39,40,41],"google-workspace","agentmail","himalaya","scrapling","osint-investigation",[43,44,45],"linux","macos","windows",{"commands":47},[48],"python","main","3e6d6b7ffb6ab43ddf3991509e5cfb5f8c90cd1a","backend\u002Foptional-skills\u002Fsecurity\u002Funbroker","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13","2026-10-13 07:57:21",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add dwikanugraha13\u002FProject-Anara --skill unbroker",{"tool":61,"command":62},"bunx","bunx skills add dwikanugraha13\u002FProject-Anara --skill unbroker",{"tool":64,"command":65},"pnpm","pnpm dlx skills add dwikanugraha13\u002FProject-Anara --skill unbroker","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002FSKILL.md",true,[69,75,81,85,89,93,97,101,105,109,113,117,121,125,129,133,137,141,145,149,153,157,161,165,169,173,177,181,185,189,193,199,203,207,211,215,219,223,227,231,235,239,243,247,251,255,259,263,267,271,275,279,283,287,291],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"assets\u002Funbroker.png","ASSET","资产文件",358001,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fassets\u002Funbroker.png",{"path":76,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":79,"sourceHtmlUrl":80},"README.md","REFERENCE","参考文档",10432,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002FREADME.md",{"path":82,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":83,"sourceHtmlUrl":84},"references\u002Fbrokers\u002Faddresses.json",1893,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Faddresses.json",{"path":86,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":87,"sourceHtmlUrl":88},"references\u002Fbrokers\u002Fadvancedbackgroundchecks.json",2927,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fadvancedbackgroundchecks.json",{"path":90,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":91,"sourceHtmlUrl":92},"references\u002Fbrokers\u002Fbeenverified.json",3622,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fbeenverified.json",{"path":94,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":95,"sourceHtmlUrl":96},"references\u002Fbrokers\u002Fclustal.json",2320,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fclustal.json",{"path":98,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":99,"sourceHtmlUrl":100},"references\u002Fbrokers\u002Fclustrmaps.json",1394,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fclustrmaps.json",{"path":102,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":103,"sourceHtmlUrl":104},"references\u002Fbrokers\u002Fcyberbackgroundchecks.json",1430,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fcyberbackgroundchecks.json",{"path":106,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":107,"sourceHtmlUrl":108},"references\u002Fbrokers\u002Ffamilytreenow.json",1484,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Ffamilytreenow.json",{"path":110,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":111,"sourceHtmlUrl":112},"references\u002Fbrokers\u002Ffastpeoplesearch.json",2275,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Ffastpeoplesearch.json",{"path":114,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":115,"sourceHtmlUrl":116},"references\u002Fbrokers\u002Fintelius.json",10909,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fintelius.json",{"path":118,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":119,"sourceHtmlUrl":120},"references\u002Fbrokers\u002Fmylife.json",974,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fmylife.json",{"path":122,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":123,"sourceHtmlUrl":124},"references\u002Fbrokers\u002Fnuwber.json",1387,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fnuwber.json",{"path":126,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":127,"sourceHtmlUrl":128},"references\u002Fbrokers\u002Fpeekyou.json",1419,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fpeekyou.json",{"path":130,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":131,"sourceHtmlUrl":132},"references\u002Fbrokers\u002Fpeoplefinders.json",1508,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fpeoplefinders.json",{"path":134,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":135,"sourceHtmlUrl":136},"references\u002Fbrokers\u002Fradaris.json",3446,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fradaris.json",{"path":138,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":139,"sourceHtmlUrl":140},"references\u002Fbrokers\u002Frehold.json",2157,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Frehold.json",{"path":142,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":143,"sourceHtmlUrl":144},"references\u002Fbrokers\u002Fsearchpeoplefree.json",1406,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fsearchpeoplefree.json",{"path":146,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":147,"sourceHtmlUrl":148},"references\u002Fbrokers\u002Fsocialcatfish.json",2130,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fsocialcatfish.json",{"path":150,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":151,"sourceHtmlUrl":152},"references\u002Fbrokers\u002Fspokeo.json",3132,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fspokeo.json",{"path":154,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":155,"sourceHtmlUrl":156},"references\u002Fbrokers\u002Fthatsthem.json",2538,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fthatsthem.json",{"path":158,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":159,"sourceHtmlUrl":160},"references\u002Fbrokers\u002Ftruepeoplesearch.json",2808,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Ftruepeoplesearch.json",{"path":162,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":163,"sourceHtmlUrl":164},"references\u002Fbrokers\u002Fusphonebook.json",1405,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fusphonebook.json",{"path":166,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":167,"sourceHtmlUrl":168},"references\u002Fbrokers\u002Fwhitepages.json",4105,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fbrokers\u002Fwhitepages.json",{"path":170,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":171,"sourceHtmlUrl":172},"references\u002Flegal\u002Fccpa.md",1125,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Flegal\u002Fccpa.md",{"path":174,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":175,"sourceHtmlUrl":176},"references\u002Flegal\u002Fdrop.md",2166,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Flegal\u002Fdrop.md",{"path":178,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":179,"sourceHtmlUrl":180},"references\u002Flegal\u002Fgdpr.md",737,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Flegal\u002Fgdpr.md",{"path":182,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":183,"sourceHtmlUrl":184},"references\u002Fmethods.md",27216,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fmethods.md",{"path":186,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":187,"sourceHtmlUrl":188},"references\u002Fsite-playbooks.md",5181,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fsite-playbooks.md",{"path":190,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":191,"sourceHtmlUrl":192},"references\u002Fstate-machine.md",3268,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Freferences\u002Fstate-machine.md",{"path":194,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":197,"sourceHtmlUrl":198},"scripts\u002Fautopilot.py","SCRIPT","脚本",21978,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fautopilot.py",{"path":200,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":201,"sourceHtmlUrl":202},"scripts\u002Fbadbool.py",6157,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fbadbool.py",{"path":204,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":205,"sourceHtmlUrl":206},"scripts\u002Fbrokers.py",2705,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fbrokers.py",{"path":208,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":209,"sourceHtmlUrl":210},"scripts\u002Fcdp.py",5911,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fcdp.py",{"path":212,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":213,"sourceHtmlUrl":214},"scripts\u002Fconfig.py",6104,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fconfig.py",{"path":216,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":217,"sourceHtmlUrl":218},"scripts\u002Fcrypto.py",3107,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fcrypto.py",{"path":220,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":221,"sourceHtmlUrl":222},"scripts\u002Fdossier.py",4963,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fdossier.py",{"path":224,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":225,"sourceHtmlUrl":226},"scripts\u002Femail_modes.py",3039,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Femail_modes.py",{"path":228,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":229,"sourceHtmlUrl":230},"scripts\u002Femailer.py",14756,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Femailer.py",{"path":232,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":233,"sourceHtmlUrl":234},"scripts\u002Fledger.py",7956,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fledger.py",{"path":236,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":237,"sourceHtmlUrl":238},"scripts\u002Flegal.py",2338,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Flegal.py",{"path":240,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":241,"sourceHtmlUrl":242},"scripts\u002Fpaths.py",2145,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fpaths.py",{"path":244,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":245,"sourceHtmlUrl":246},"scripts\u002Fpdd.py",47203,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fpdd.py",{"path":248,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":249,"sourceHtmlUrl":250},"scripts\u002Fregistry.py",13826,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fregistry.py",{"path":252,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":253,"sourceHtmlUrl":254},"scripts\u002Freport.py",7189,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Freport.py",{"path":256,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":257,"sourceHtmlUrl":258},"scripts\u002Fscan.py",1308,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fscan.py",{"path":260,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":261,"sourceHtmlUrl":262},"scripts\u002Fstorage.py",4453,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fstorage.py",{"path":264,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":265,"sourceHtmlUrl":266},"scripts\u002Ftiers.py",14182,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Ftiers.py",{"path":268,"resourceType":195,"resourceTypeDesc":196,"sizeBytes":269,"sourceHtmlUrl":270},"scripts\u002Fvectors.py",2122,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Fscripts\u002Fvectors.py",{"path":272,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":273,"sourceHtmlUrl":274},"templates\u002Fconsent\u002Fauthorization.md",742,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Ftemplates\u002Fconsent\u002Fauthorization.md",{"path":276,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":277,"sourceHtmlUrl":278},"templates\u002Femails\u002Fccpa-authorized-agent.txt",792,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Ftemplates\u002Femails\u002Fccpa-authorized-agent.txt",{"path":280,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":281,"sourceHtmlUrl":282},"templates\u002Femails\u002Fccpa-deletion.txt",686,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Ftemplates\u002Femails\u002Fccpa-deletion.txt",{"path":284,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":285,"sourceHtmlUrl":286},"templates\u002Femails\u002Fccpa-indirect-deletion.txt",1472,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Ftemplates\u002Femails\u002Fccpa-indirect-deletion.txt",{"path":288,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":289,"sourceHtmlUrl":290},"templates\u002Femails\u002Fgdpr-erasure.txt",606,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Ftemplates\u002Femails\u002Fgdpr-erasure.txt",{"path":292,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":293,"sourceHtmlUrl":294},"templates\u002Femails\u002Fgeneric-optout.txt",470,"https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Foptional-skills\u002Fsecurity\u002Funbroker\u002Ftemplates\u002Femails\u002Fgeneric-optout.txt",[296,303,310,318,325,332],{"skillNo":297,"name":298,"nameZh":6,"description":299,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":300,"categories":301,"keywords":302,"publishedAt":18},"01M7CVK8W5MPQ8H45A6XE6RPK1","youtube-content","YouTube transcripts to summaries, threads, blogs.","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Fyoutube-content\u002FSKILL.md",[],[],{"skillNo":304,"name":305,"nameZh":6,"description":306,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":307,"categories":308,"keywords":309,"publishedAt":18},"01M7CVK56KG9399ZZVE9KM54WS","xurl","X\u002FTwitter via xurl CLI: raw post search, posting, DM, media.","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Fxurl\u002FSKILL.md",[],[],{"skillNo":311,"name":312,"nameZh":6,"description":313,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":314,"sourceHtmlUrl":315,"categories":316,"keywords":317,"publishedAt":18},"01M7CVK22A44A88NPPBAC1FVQP","xlsx","Create, read, edit Excel .xlsx workbooks and CSVs.","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Fxlsx\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Fxlsx\u002FSKILL.md",[],[],{"skillNo":319,"name":320,"nameZh":6,"description":321,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":322,"categories":323,"keywords":324,"publishedAt":18},"01M7CVJX88C8PY2T6RQZ5GCG15","weights-and-biases","W&B: log ML experiments, sweeps, model registry, dashboards.","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Fweights-and-biases\u002FSKILL.md",[],[],{"skillNo":326,"name":327,"nameZh":6,"description":328,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":329,"categories":330,"keywords":331,"publishedAt":18},"01M7CVJR3W1FBRGC56DG6BJ1Y3","weekly-review-planning","Weekly reset: commitments, stalled work, next-week plan.","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Fweekly-review-planning\u002FSKILL.md",[],[],{"skillNo":333,"name":334,"nameZh":6,"description":335,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":336,"categories":337,"keywords":338,"publishedAt":18},"01M7CVJGPJNPQDM8S1KTTMJN7M","touchdesigner-mcp","Control TouchDesigner via twozero MCP.","https:\u002F\u002Fgithub.com\u002Fdwikanugraha13\u002FProject-Anara\u002Fblob\u002Fmain\u002Fbackend\u002Fskills\u002Ftouchdesigner-mcp\u002FSKILL.md",[],[]]