[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M75TSF64Z2R966N19QSVR34Q":3,"skill-same:01M75TSF64Z2R966N19QSVR34Q":45},{"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":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":28,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M75TSF64Z2R966N19QSVR34Q","luna-ingest",null,"Use when saving GitHub repos or issues, or bitbucket.org repos, issues or PRs, as structured luna notes.","yotamleo-org\u002FHimmel","yotamleo-org","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F344898079?v=4","01M75TKGRE629BDM7G461ABVZP","01M75TKGS2GVCFR1A28MBMK997",15,"MIT","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fmarketplace\u002Fplugins\u002Fobsidian-triage\u002Fskills\u002Fluna-ingest\u002FSKILL.md",[],[],"2026-11-12 13:55:43",false,"---\nname: luna-ingest\ndescription: Use when saving GitHub repos or issues, or bitbucket.org repos, issues or PRs, as structured luna notes.\n---\n\n# luna-ingest — chain-following triage\n\nTake a github OR bitbucket.org URL, follow its README references one hop (github sources only — see below), classify each, and produce a structured tech-ref note in the luna vault.\n\n**Scope:** github + bitbucket.org URLs. Twitter (`x.com\u002F...`) and article URLs raise an explicit \"not yet\" error.\n- **github.com:** repo URLs run Phases 1–4; issue URLs (`\u002Fissues\u002F\u003Cn>`) take the issue branch; PR (`\u002Fpull\u002F\u003Cn>`) and discussion URLs remain out of scope.\n- **bitbucket.org:** repo, PR (`\u002Fpull-requests\u002F\u003Cn>`), and issue (`\u002Fissues\u002F\u003Cn>`) URLs take the **Bitbucket branch** below. They dispatch to the himmel bitbucket CLI (`node scripts\u002Fbitbucket\u002Fdist\u002Findex.js …`) instead of `gh`.\n\n**Forge routing:** the URL **host** selects the path — parallel to how the dev-loop routes on the git origin. `github.com` (or bare `\u003Cowner>\u002F\u003Crepo>`) → the github Phases below, unchanged. `bitbucket.org` → the **Bitbucket branch**. Everything in Phases 1–4 and the issue branch is the github path; the Bitbucket branch reuses the shared Phase 1.5 safety pre-filter and Phase 3\u002F4 heuristics but swaps the fetch transport.\n\n**Invocation surfaces:**\n\n- **User prompt:** `\u002Fluna-ingest \u003Cgithub-url> [--vault \u003Cpath>] [--dest \u003Ccategory>] [--limit \u003CN>] [--deep] [--research] [--dry-run]` — handled by the thin wrapper at `.claude\u002Fcommands\u002Fluna-ingest.md` which delegates here.\n- **Programmatic Skill-tool dispatch (no headless claude):** `Skill { skill: \"obsidian-triage:luna-ingest\", args: \"\u003Cgithub-url> [\u003Cflags...>]\" }` from inside another runbook (e.g. `\u002Fharvest-clips` github-URL branch).\n\nBoth surfaces feed the same argument shape into the runbook below. Wherever this document references `$ARGUMENTS`, treat it as the literal arg string supplied via either invocation path.\n\n## Inputs\n\n`$ARGUMENTS` is `\u003Curl> [--vault \u003Cpath>] [--dest \u003Ccategory>] [--limit \u003CN>] [--dry-run]`.\n\nParse:\n- `\u003Curl>` — required. Accepted shapes:\n  - `https:\u002F\u002Fgithub.com\u002F\u003Cowner>\u002F\u003Crepo>` (+ `\u002F`, `\u002Ftree\u002F\u003Cbranch>`, `\u002Fblob\u002F\u003Cbranch>\u002F\u003Cpath>`)\n  - `https:\u002F\u002Fgithub.com\u002F\u003Cowner>\u002F\u003Crepo>\u002Fissues\u002F\u003Cn>` — routes to the github issue branch\n  - bare `\u003Cowner>\u002F\u003Crepo>` — treated as github\n  - `https:\u002F\u002Fbitbucket.org\u002F\u003Cws>\u002F\u003Crepo>` (+ `\u002F`, `\u002Fsrc\u002F\u003Cbranch>\u002F...`) — routes to the **Bitbucket branch**\n  - `https:\u002F\u002Fbitbucket.org\u002F\u003Cws>\u002F\u003Crepo>\u002Fpull-requests\u002F\u003Cn>` — Bitbucket PR branch\n  - `https:\u002F\u002Fbitbucket.org\u002F\u003Cws>\u002F\u003Crepo>\u002Fissues\u002F\u003Cn>` — Bitbucket issue branch\n- `--vault \u003Cpath>` — luna vault root. Default: `$HOME\u002FDocuments\u002Fluna`. Refuse with rc=2 if the dir does not exist OR is not an Obsidian vault (no `.obsidian\u002F` dir).\n- `--dest \u003Ccategory>` — vault destination category override. Default: `30-Resources\u002FTech`. Validate against the vault's PARA folders (`00-Inbox`, `10-Projects\u002F*`, `20-Areas\u002F*`, `30-Resources\u002FTech`, `30-Resources\u002FConcepts`, `30-Resources\u002FComponents`). Anything else → rc=2.\n- `--limit \u003CN>` — max 1-hop refs to follow. Default `250` (capture all refs by default — the prior `[1,25]` clamp left high-fanout awesome-lists partial and silently dropped sources). No upper clamp; reject `\u003C 1` with rc=1. `250` is a generous safety ceiling for unattended `harvest-clips` dispatch (which passes no `--limit`), not a quality cap — the largest list we have hit is 188 refs. Pass `--limit \u003CN>` to bound a noisy meta-list downward. The rate-limit contract (api_failure verdict + rc=4 resume) is the backstop if a pathological list exceeds the gh-api hourly budget.\n- `--dry-run` — print what would be written + verdicts, touch no files. rc=0.\n- `--allow-unsafe` — opt-in flag to ingest a repo flagged by the safety-keyword pre-filter (Phase 1.5). Without this flag, repos whose name\u002Fdescription\u002Ftopics match red-flag terms exit rc=6 before fetching the README.\n- `--repo \u003Cws\u002Frepo>` is NOT a luna-ingest flag — it is the bitbucket CLI's own override, set internally by the Bitbucket branch from the parsed URL. Operators pass the bitbucket.org URL, not `--repo`.\n- `--deep` — after classification, run the component scan\n  (`tools\u002Fcomponent-scan.mjs`) over the source repo (always) and over each\n  1-hop ref verdicted `integrate` or `take-parts`. Inventories reusable\n  components (skills\u002Fcommands\u002Fagents\u002Ftools\u002Fplugin manifests) into a\n  `## Reusable components` section of the Tech note AND a cross-repo-deduped\n  `30-Resources\u002FComponents\u002F` library. Off by default (MVP keeps the light path).\n- `--components-dir \u003Crel>` — vault-relative Components library root.\n  Default `30-Resources\u002FComponents`. Only meaningful with `--deep`.\n- `--max-components \u003CN>` — per-repo component cap (tail-skip beyond). Default 60.\n- `--skill-scan` — run SkillSpector Stage 1 (static, `--no-llm`, no egress) over each skill directory in the source repo (Phase 3.7) and store the JSON reports beside the note. Advisory only, never gates. Off by default.\n- `--research` — after classification, web-research the SOURCE repo\n  (maturity\u002Fadoption, notable alternatives, known limitations, community\n  sentiment, recency) using Claude's own WebSearch\u002FWebFetch tools and inject a\n  `## Research enrichment` section into the Tech note. Off by default.\n  **Orthogonal to `--deep`** — they enrich different axes (`--deep` = reusable\n  components inventoried from the code; `--research` = external context about\n  the repo) and may be combined (`--deep --research` runs both). This is the\n  lightweight inline path: NO external Python (`research_deep.py`), NO\n  Perplexity\u002FGrok keys, NO vault-wide propagation — those belong to the full\n  `\u002Fresearch-deep` command, deferred. Source repo\n  only in MVP (1-hop refs are not researched). The harvest-clips github\n  dispatch MUST NOT pass `--research` (cost + scope safety).\n\nReject `x.com\u002F`, `twitter.com\u002F`, or any host that is neither `github.com` nor `bitbucket.org` with:\n\n```\nERR luna-ingest: only github.com and bitbucket.org URLs supported. Twitter + article inputs are not supported yet.\n```\n\n## Issue branch — github issue URLs\n\n**Read on demand:** When the URL is a github issue URL (`\u002Fissues\u002F\u003Cn>`) — before doing anything else for it. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fissue-branch.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Bitbucket branch — bitbucket.org URLs\n\n**Read on demand:** When the URL host is `bitbucket.org` (repo, PR or issue) — before doing anything else for it. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fbitbucket-branch.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 1 — Source fetch (github URL)\n**Read on demand:** Whenever the source is a github repo URL (every github run); before fetching. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-1-source-fetch.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 1.5 — Safety pre-filter\n**Read on demand:** After Phase 1, before Phase 2, on every github run; the matching algorithm and rc=6 behaviour live there. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-1-5-safety-prefilter.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 1.6 — Advisory repo security scan\n**Read on demand:** After Phase 1.5, before Phase 2, on every github run; the command, finding kinds and the note fields live there. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-1-6-repo-scan.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 2 — Reference extraction\n**Read on demand:** When extracting refs from the README (Phase 2) and applying the `--limit` cap and the README-size warning. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-2-reference-extraction.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3 — Per-ref classification\n**Read on demand:** When classifying each ref (Phase 3): the failure classes, the verdict heuristics and the confidence floor. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-classification.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3.5 — Deep component scan (`--deep` only)\n\n**Read on demand:** Only when `--deep` was passed; skip otherwise. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-5-deep-scan.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3.6 — Research enrichment (`--research` only)\n\n**Read on demand:** Only when `--research` was passed; skip otherwise. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-6-research.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3.7 — Advisory skill scan (`--skill-scan` only)\n\n**Read on demand:** Only when `--skill-scan` was passed; skip otherwise. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-7-skill-scan.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 4 — Synthesis\n\nCompute a destination path under `\u003Cvault>\u002F\u003Cdest-category>\u002F\u003Crepo-slug>.md` where:\n- `\u003Crepo-slug>` = lowercase, hyphens, max 60 chars. e.g. `kepano\u002Fobsidian-skills` → `kepano-obsidian-skills`. Slugify: replace any non-`[a-z0-9-]` with `-`, collapse repeats, strip leading\u002Ftrailing `-`.\n- If `\u003Cdest-category>` was passed via `--dest`, use it verbatim. Else default `30-Resources\u002FTech`.\n\n**Path safety invariant:** before any Write, canonicalise the resolved destination via `realpath -m` and verify it starts with the canonicalised vault root. Refuse with rc=2 if a `..` traversal or symlink escape would land outside the vault. This guards against malformed `--dest` or weirdly-named refs.\n\nIf the target file already exists: refuse with rc=3 unless `--force` was passed (NOT supported in MVP — rc=3 with a hint to delete + retry).\n\n### Trust-tier auto-classification\n\n**Read on demand:** Always, in Phase 4 before the Write step, to compute `trust_tier` and `trust_tier_reason` (the Bitbucket branch's B-repo-3 deltas and the issue branch's I-3 also apply these rules). Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Ftrust-tier.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n### Tags derivation\n\n**Read on demand:** Always, in Phase 4 before building the frontmatter (every emitted note needs `tags:`). Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Ftags-derivation.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n```markdown\n---\ntype: tech-ingest\nsource: \u003CURL>\nsource_type: github\ningested_at: \u003CUTC ISO-8601>\ningest_spec_version: 1        # version of OUR OWN extraction rules, independent of upstream movement. Bump this when the extraction rules change so old notes are known-truncated\u002Fknown-stale on our side even if upstream hasn't moved (e.g. notes written before the --limit\u002Ftail-skip work are truncated and nothing else says so).\ningest_scope:                 # machine-readable, not a line in the body\n  limit: \u003Ceffective --limit used for this ingest>\n  tail_skipped: \u003Ccount, 0 if none>\nupstream_commit: \u003Cdefault-branch HEAD SHA at fetch time, or omit on fetch failure>\nupstream_pushed_at: \u003Crepo pushed_at, ISO-8601>   # parseable duplicate of the value also folded into trust_tier_reason prose below\nreadme_sha256: \u003Csha256 of the decoded README bytes, or omit if README missing>   # replaces README-byte-count as the change-detection proxy\n# last_revalidated is OMITTED here — never default it to ingested_at. \"Never\n# checked\" and \"checked, unchanged\" must not write the same value. It is written ONLY when a\n# real revalidation sweep runs later (see \"Revalidation invariant\" below), paired with a\n# revalidation_delta line.\ningest_command: \u002Fluna-ingest\nverdict_summary:\n  integrate: \u003Ccount>\n  take-parts: \u003Ccount>\n  inspire: \u003Ccount>\n  skip: \u003Ccount>\n  api_failure: \u003Ccount>   # refs we couldn't classify due to rate-limit \u002F network \u002F auth. NON-ZERO means partial run.\nlicense: \u003CSPDX or unknown>\nstars: \u003CN>\nlanguage: \u003Cprimary>\ntopics:\n  - \u003Ctopic>\ntags:                       # REQUIRED — never omit. Derivation rules in \"Tags derivation\" above.\n  - github\n  - \u003Ctag>\ntrust_tier: \u003Ctier>          # anthropic-official | known-author | community-active | community-thin | unknown-risk\ntrust_tier_reason: \u003Ctext>   # 1-line justification (e.g. \"owner=anthropics → anthropic-official\"; \"owner=kepano in allowlist → known-author\"; \"stargazers_count=12, pushed_at=2025-04-01 → community-thin\"). Field name is `owner=` (the lowercased value extracted in trust-tier rule logic), matching the trust-tier rule wording — NOT `owner_org=`.\nsafety_flag: \u003Cterm-or-blank> # blank if Phase 1.5 found no red-flag terms; matched term otherwise. NON-BLANK forces trust_tier=unknown-risk via rule 0 short-circuit in trust-tier classification.\nrepo_scan: \u003Cclean|flagged|unavailable>   # Phase 1.6 advisory scan result; never gates trust_tier.\nskill_scan: \u003Cclean|flagged|unavailable>  # set ONLY when --skill-scan ran (Phase 3.7); omit otherwise. Advisory; never gates trust_tier.\nresearch_enriched: \u003Ctrue-or-omit>  # set true ONLY when --research ran (Phase 3.6) AND produced real findings. Omit the field entirely when --research was not passed OR enrichment degraded (dry-run short-circuit \u002F web-failure placeholder) — a degraded note must not advertise itself as enriched to downstream readers (\u002Fsynthesize-clips, revalidation tooling).\n---\n\n# \u003Crepo-name>\n\n## Source\n\n[\u003Cfull_name>](\u003CURL>) — \u003Cdescription from gh api>\n\n- **Language:** \u003Clanguage> | **Stars:** \u003CN> | **License:** \u003CSPDX>\n- **Last push:** \u003Cpushed_at>\n- **Topics:** \u003Ctopics joined by \", \">\n\n## Summary\n\n\u003C3-5 bullet TL;DR derived from README + repo metadata>\n\n## Referenced repos (1-hop)\n\n\u003Cfor each ref with confidence ≥0.6, sorted by verdict severity integrate > take-parts > inspire > skip>\n\n### \u003Cref-full-name> — \u003Cverdict>\n\n- URL: https:\u002F\u002Fgithub.com\u002F\u003Cref>\n- Language: \u003Clang> | Stars: \u003CN> | License: \u003CSPDX>\n- **Rationale:** \u003C1-3 sentences>\n\u003Cif verdict == take-parts>\n- **What to lift:** \u003Cfile\u002Fpath\u002Farea to inspect first>\n\u003C\u002Fif>\n\n## Security scan (advisory)\n\n\u003CPhase 1.6: the scan `summary` line, then one bullet per finding. Always present on a github run.>\n\n## Skill scan (advisory)\n\n\u003COnly present when --skill-scan was passed (Phase 3.7): one line per scanned skill, as the phase reference specifies. If --skill-scan was NOT passed, omit this section entirely.>\n\n## Research enrichment\n\n\u003COnly present when --research was passed (Phase 3.6). 4–8 bullets synthesizing\nthe WebSearch findings about the SOURCE repo — maturity\u002Fadoption, notable\nalternatives, known limitations, community sentiment. Each external claim ends\nwith a recency marker + source URL inline, e.g.:\n`- Actively maintained; v2.3 released as of Mar 2026 ([release notes](https:\u002F\u002F…)).`\nOn dry-run, render `_(dry-run — research not executed)_`. On web failure,\nrender `_(research enrichment requested but web search returned no usable\nresults)_`. If --research was NOT passed, omit this section entirely.>\n\n## Reusable components\n\n\u003COnly present when --deep was passed. Compose from the source-repo scan\nJSON. For each component in `components[]`, one bullet:\n`- **\u003Cname>** (\u003Ctype>) — \u003Cdescription>. _Reuse:_ \u003C1-2 line how-to-rephrase\nfor himmel\u002Fluna — YOUR judgment from the description + type>.`\nThen, for each integrate\u002Ftake-parts ref that was scanned, one summary line:\n`- _Ref \u003Co>\u002F\u003Cr>: N components inventoried → see [[30-Resources\u002FComponents\u002F]]._`\nIf --deep was NOT passed, omit this section entirely.>\n\n## Weak refs (low-confidence, appendix)\n\n\u003Cfor each ref with confidence \u003C0.6 — same structure, condensed>\n\n## Unclassified (transient — re-run to retry)\n\n\u003Cfor each ref with `verdict: api_failure` — list URL + verdict_reason. surfaces rate-limited \u002F network-failed refs distinctly from § Skip so operator knows what to retry. Empty section reads `_(none — all refs classified)_`.>\n\n## Action items\n\n\u003Cextracted from README headings like \"TODO\", \"Roadmap\", \"Known issues\" — convert each to `- [ ] \u003Citem>`. If none found: `_(none — repo README has no TODO\u002FRoadmap section)_`>\n\n## Open questions\n\n\u003Cgaps the chain-follow could not resolve. Always include: \"Verify integrate-verdict refs by reading their root README before taking a dep on any of them.\">\n\n## Audit log\n\n- Refs scanned: \u003CN>\n- Refs above confidence floor (0.6): \u003CK>\n- Refs marked api_failure: \u003CF> (if >0: rc=4; re-run to retry classification)\n- Tail-skipped: \u003CT> (set if more refs existed than --limit)\n- API calls: \u003CN+2> (+1 over the prior count for the `upstream_commit` fetch)\n- README size: \u003Cbytes> (warn if >512 KB; --limit caps API-call count, not text size)\n- Trust tier: \u003Ctier> (\u003Creason>)\n- Safety flag: \u003Cterm-or-none>   (non-blank value reached rc=6 unless --allow-unsafe; document the term in this row regardless)\n```\n\n### Revalidation invariant\n\n**Read on demand:** Only when writing or reviewing a `## Re-validation` section; a fresh ingest never does. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Frevalidation-invariant.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n### Semantic layer (schema defined, NOT yet populated)\n\n**Read on demand:** Only when touching the not-yet-populated semantic fields (`concepts`, `solves`, `maps_to`, `value_tier`, `value_reason`); this MVP never writes them. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fsemantic-layer.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 5 — Daily-note backref (deferred to v2)\n\nMVP does NOT touch the daily note. v2 picks up the obsidian-triage Phase 5 dedup-by-backreference contract.\n\n## Phase 6 — Promotion candidate (deferred to v2)\n\nMVP does NOT write `00-Inbox\u002Fpromote-\u003Cslug>.md`. v2 adds this when any verdict ∈ {integrate, take-parts}.\n\n## Dry-run mode\n\nWhen `--dry-run` is passed:\n- Print \"DRY luna-ingest: would write \u003Cfull-path>\"\n- Print the synthesized markdown body to stdout, prefixed with 4 spaces\n- Print per-ref verdict summary table\n- With `--research`: skip the Phase 3.6 web queries (print `DRY luna-ingest:\n  would run research enrichment for \u003Cowner_repo>`); the `## Research\n  enrichment` section renders `_(dry-run — research not executed)_`.\n- rc=0, touch no files in the vault\n\n## Exit codes\n\n**Read on demand:** Before choosing or interpreting an exit code (0 ok, 1 usage, 2 env, 3 target exists, 4 partial, 5 source fetch failed, 6 safety-blocked). Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fexit-codes.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Out of scope (v2 follow-ups)\n\n**Read on demand:** Only when asked whether an input form or feature is supported. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fout-of-scope.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n","# luna-ingest — chain-following triage\n\nTake a github OR bitbucket.org URL, follow its README references one hop (github sources only — see below), classify each, and produce a structured tech-ref note in the luna vault.\n\n**Scope:** github + bitbucket.org URLs. Twitter (`x.com\u002F...`) and article URLs raise an explicit \"not yet\" error.\n- **github.com:** repo URLs run Phases 1–4; issue URLs (`\u002Fissues\u002F\u003Cn>`) take the issue branch; PR (`\u002Fpull\u002F\u003Cn>`) and discussion URLs remain out of scope.\n- **bitbucket.org:** repo, PR (`\u002Fpull-requests\u002F\u003Cn>`), and issue (`\u002Fissues\u002F\u003Cn>`) URLs take the **Bitbucket branch** below. They dispatch to the himmel bitbucket CLI (`node scripts\u002Fbitbucket\u002Fdist\u002Findex.js …`) instead of `gh`.\n\n**Forge routing:** the URL **host** selects the path — parallel to how the dev-loop routes on the git origin. `github.com` (or bare `\u003Cowner>\u002F\u003Crepo>`) → the github Phases below, unchanged. `bitbucket.org` → the **Bitbucket branch**. Everything in Phases 1–4 and the issue branch is the github path; the Bitbucket branch reuses the shared Phase 1.5 safety pre-filter and Phase 3\u002F4 heuristics but swaps the fetch transport.\n\n**Invocation surfaces:**\n\n- **User prompt:** `\u002Fluna-ingest \u003Cgithub-url> [--vault \u003Cpath>] [--dest \u003Ccategory>] [--limit \u003CN>] [--deep] [--research] [--dry-run]` — handled by the thin wrapper at `.claude\u002Fcommands\u002Fluna-ingest.md` which delegates here.\n- **Programmatic Skill-tool dispatch (no headless claude):** `Skill { skill: \"obsidian-triage:luna-ingest\", args: \"\u003Cgithub-url> [\u003Cflags...>]\" }` from inside another runbook (e.g. `\u002Fharvest-clips` github-URL branch).\n\nBoth surfaces feed the same argument shape into the runbook below. Wherever this document references `$ARGUMENTS`, treat it as the literal arg string supplied via either invocation path.\n\n## Inputs\n\n`$ARGUMENTS` is `\u003Curl> [--vault \u003Cpath>] [--dest \u003Ccategory>] [--limit \u003CN>] [--dry-run]`.\n\nParse:\n- `\u003Curl>` — required. Accepted shapes:\n  - `https:\u002F\u002Fgithub.com\u002F\u003Cowner>\u002F\u003Crepo>` (+ `\u002F`, `\u002Ftree\u002F\u003Cbranch>`, `\u002Fblob\u002F\u003Cbranch>\u002F\u003Cpath>`)\n  - `https:\u002F\u002Fgithub.com\u002F\u003Cowner>\u002F\u003Crepo>\u002Fissues\u002F\u003Cn>` — routes to the github issue branch\n  - bare `\u003Cowner>\u002F\u003Crepo>` — treated as github\n  - `https:\u002F\u002Fbitbucket.org\u002F\u003Cws>\u002F\u003Crepo>` (+ `\u002F`, `\u002Fsrc\u002F\u003Cbranch>\u002F...`) — routes to the **Bitbucket branch**\n  - `https:\u002F\u002Fbitbucket.org\u002F\u003Cws>\u002F\u003Crepo>\u002Fpull-requests\u002F\u003Cn>` — Bitbucket PR branch\n  - `https:\u002F\u002Fbitbucket.org\u002F\u003Cws>\u002F\u003Crepo>\u002Fissues\u002F\u003Cn>` — Bitbucket issue branch\n- `--vault \u003Cpath>` — luna vault root. Default: `$HOME\u002FDocuments\u002Fluna`. Refuse with rc=2 if the dir does not exist OR is not an Obsidian vault (no `.obsidian\u002F` dir).\n- `--dest \u003Ccategory>` — vault destination category override. Default: `30-Resources\u002FTech`. Validate against the vault's PARA folders (`00-Inbox`, `10-Projects\u002F*`, `20-Areas\u002F*`, `30-Resources\u002FTech`, `30-Resources\u002FConcepts`, `30-Resources\u002FComponents`). Anything else → rc=2.\n- `--limit \u003CN>` — max 1-hop refs to follow. Default `250` (capture all refs by default — the prior `[1,25]` clamp left high-fanout awesome-lists partial and silently dropped sources). No upper clamp; reject `\u003C 1` with rc=1. `250` is a generous safety ceiling for unattended `harvest-clips` dispatch (which passes no `--limit`), not a quality cap — the largest list we have hit is 188 refs. Pass `--limit \u003CN>` to bound a noisy meta-list downward. The rate-limit contract (api_failure verdict + rc=4 resume) is the backstop if a pathological list exceeds the gh-api hourly budget.\n- `--dry-run` — print what would be written + verdicts, touch no files. rc=0.\n- `--allow-unsafe` — opt-in flag to ingest a repo flagged by the safety-keyword pre-filter (Phase 1.5). Without this flag, repos whose name\u002Fdescription\u002Ftopics match red-flag terms exit rc=6 before fetching the README.\n- `--repo \u003Cws\u002Frepo>` is NOT a luna-ingest flag — it is the bitbucket CLI's own override, set internally by the Bitbucket branch from the parsed URL. Operators pass the bitbucket.org URL, not `--repo`.\n- `--deep` — after classification, run the component scan\n  (`tools\u002Fcomponent-scan.mjs`) over the source repo (always) and over each\n  1-hop ref verdicted `integrate` or `take-parts`. Inventories reusable\n  components (skills\u002Fcommands\u002Fagents\u002Ftools\u002Fplugin manifests) into a\n  `## Reusable components` section of the Tech note AND a cross-repo-deduped\n  `30-Resources\u002FComponents\u002F` library. Off by default (MVP keeps the light path).\n- `--components-dir \u003Crel>` — vault-relative Components library root.\n  Default `30-Resources\u002FComponents`. Only meaningful with `--deep`.\n- `--max-components \u003CN>` — per-repo component cap (tail-skip beyond). Default 60.\n- `--skill-scan` — run SkillSpector Stage 1 (static, `--no-llm`, no egress) over each skill directory in the source repo (Phase 3.7) and store the JSON reports beside the note. Advisory only, never gates. Off by default.\n- `--research` — after classification, web-research the SOURCE repo\n  (maturity\u002Fadoption, notable alternatives, known limitations, community\n  sentiment, recency) using Claude's own WebSearch\u002FWebFetch tools and inject a\n  `## Research enrichment` section into the Tech note. Off by default.\n  **Orthogonal to `--deep`** — they enrich different axes (`--deep` = reusable\n  components inventoried from the code; `--research` = external context about\n  the repo) and may be combined (`--deep --research` runs both). This is the\n  lightweight inline path: NO external Python (`research_deep.py`), NO\n  Perplexity\u002FGrok keys, NO vault-wide propagation — those belong to the full\n  `\u002Fresearch-deep` command, deferred. Source repo\n  only in MVP (1-hop refs are not researched). The harvest-clips github\n  dispatch MUST NOT pass `--research` (cost + scope safety).\n\nReject `x.com\u002F`, `twitter.com\u002F`, or any host that is neither `github.com` nor `bitbucket.org` with:\n\n```\nERR luna-ingest: only github.com and bitbucket.org URLs supported. Twitter + article inputs are not supported yet.\n```\n\n## Issue branch — github issue URLs\n\n**Read on demand:** When the URL is a github issue URL (`\u002Fissues\u002F\u003Cn>`) — before doing anything else for it. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fissue-branch.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Bitbucket branch — bitbucket.org URLs\n\n**Read on demand:** When the URL host is `bitbucket.org` (repo, PR or issue) — before doing anything else for it. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fbitbucket-branch.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 1 — Source fetch (github URL)\n**Read on demand:** Whenever the source is a github repo URL (every github run); before fetching. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-1-source-fetch.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 1.5 — Safety pre-filter\n**Read on demand:** After Phase 1, before Phase 2, on every github run; the matching algorithm and rc=6 behaviour live there. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-1-5-safety-prefilter.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 1.6 — Advisory repo security scan\n**Read on demand:** After Phase 1.5, before Phase 2, on every github run; the command, finding kinds and the note fields live there. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-1-6-repo-scan.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 2 — Reference extraction\n**Read on demand:** When extracting refs from the README (Phase 2) and applying the `--limit` cap and the README-size warning. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-2-reference-extraction.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3 — Per-ref classification\n**Read on demand:** When classifying each ref (Phase 3): the failure classes, the verdict heuristics and the confidence floor. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-classification.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3.5 — Deep component scan (`--deep` only)\n\n**Read on demand:** Only when `--deep` was passed; skip otherwise. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-5-deep-scan.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3.6 — Research enrichment (`--research` only)\n\n**Read on demand:** Only when `--research` was passed; skip otherwise. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-6-research.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 3.7 — Advisory skill scan (`--skill-scan` only)\n\n**Read on demand:** Only when `--skill-scan` was passed; skip otherwise. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fphase-3-7-skill-scan.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 4 — Synthesis\n\nCompute a destination path under `\u003Cvault>\u002F\u003Cdest-category>\u002F\u003Crepo-slug>.md` where:\n- `\u003Crepo-slug>` = lowercase, hyphens, max 60 chars. e.g. `kepano\u002Fobsidian-skills` → `kepano-obsidian-skills`. Slugify: replace any non-`[a-z0-9-]` with `-`, collapse repeats, strip leading\u002Ftrailing `-`.\n- If `\u003Cdest-category>` was passed via `--dest`, use it verbatim. Else default `30-Resources\u002FTech`.\n\n**Path safety invariant:** before any Write, canonicalise the resolved destination via `realpath -m` and verify it starts with the canonicalised vault root. Refuse with rc=2 if a `..` traversal or symlink escape would land outside the vault. This guards against malformed `--dest` or weirdly-named refs.\n\nIf the target file already exists: refuse with rc=3 unless `--force` was passed (NOT supported in MVP — rc=3 with a hint to delete + retry).\n\n### Trust-tier auto-classification\n\n**Read on demand:** Always, in Phase 4 before the Write step, to compute `trust_tier` and `trust_tier_reason` (the Bitbucket branch's B-repo-3 deltas and the issue branch's I-3 also apply these rules). Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Ftrust-tier.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n### Tags derivation\n\n**Read on demand:** Always, in Phase 4 before building the frontmatter (every emitted note needs `tags:`). Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Ftags-derivation.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n```markdown\n---\ntype: tech-ingest\nsource: \u003CURL>\nsource_type: github\ningested_at: \u003CUTC ISO-8601>\ningest_spec_version: 1        # version of OUR OWN extraction rules, independent of upstream movement. Bump this when the extraction rules change so old notes are known-truncated\u002Fknown-stale on our side even if upstream hasn't moved (e.g. notes written before the --limit\u002Ftail-skip work are truncated and nothing else says so).\ningest_scope:                 # machine-readable, not a line in the body\n  limit: \u003Ceffective --limit used for this ingest>\n  tail_skipped: \u003Ccount, 0 if none>\nupstream_commit: \u003Cdefault-branch HEAD SHA at fetch time, or omit on fetch failure>\nupstream_pushed_at: \u003Crepo pushed_at, ISO-8601>   # parseable duplicate of the value also folded into trust_tier_reason prose below\nreadme_sha256: \u003Csha256 of the decoded README bytes, or omit if README missing>   # replaces README-byte-count as the change-detection proxy\n# last_revalidated is OMITTED here — never default it to ingested_at. \"Never\n# checked\" and \"checked, unchanged\" must not write the same value. It is written ONLY when a\n# real revalidation sweep runs later (see \"Revalidation invariant\" below), paired with a\n# revalidation_delta line.\ningest_command: \u002Fluna-ingest\nverdict_summary:\n  integrate: \u003Ccount>\n  take-parts: \u003Ccount>\n  inspire: \u003Ccount>\n  skip: \u003Ccount>\n  api_failure: \u003Ccount>   # refs we couldn't classify due to rate-limit \u002F network \u002F auth. NON-ZERO means partial run.\nlicense: \u003CSPDX or unknown>\nstars: \u003CN>\nlanguage: \u003Cprimary>\ntopics:\n  - \u003Ctopic>\ntags:                       # REQUIRED — never omit. Derivation rules in \"Tags derivation\" above.\n  - github\n  - \u003Ctag>\ntrust_tier: \u003Ctier>          # anthropic-official | known-author | community-active | community-thin | unknown-risk\ntrust_tier_reason: \u003Ctext>   # 1-line justification (e.g. \"owner=anthropics → anthropic-official\"; \"owner=kepano in allowlist → known-author\"; \"stargazers_count=12, pushed_at=2025-04-01 → community-thin\"). Field name is `owner=` (the lowercased value extracted in trust-tier rule logic), matching the trust-tier rule wording — NOT `owner_org=`.\nsafety_flag: \u003Cterm-or-blank> # blank if Phase 1.5 found no red-flag terms; matched term otherwise. NON-BLANK forces trust_tier=unknown-risk via rule 0 short-circuit in trust-tier classification.\nrepo_scan: \u003Cclean|flagged|unavailable>   # Phase 1.6 advisory scan result; never gates trust_tier.\nskill_scan: \u003Cclean|flagged|unavailable>  # set ONLY when --skill-scan ran (Phase 3.7); omit otherwise. Advisory; never gates trust_tier.\nresearch_enriched: \u003Ctrue-or-omit>  # set true ONLY when --research ran (Phase 3.6) AND produced real findings. Omit the field entirely when --research was not passed OR enrichment degraded (dry-run short-circuit \u002F web-failure placeholder) — a degraded note must not advertise itself as enriched to downstream readers (\u002Fsynthesize-clips, revalidation tooling).\n---\n\n# \u003Crepo-name>\n\n## Source\n\n[\u003Cfull_name>](\u003CURL>) — \u003Cdescription from gh api>\n\n- **Language:** \u003Clanguage> | **Stars:** \u003CN> | **License:** \u003CSPDX>\n- **Last push:** \u003Cpushed_at>\n- **Topics:** \u003Ctopics joined by \", \">\n\n## Summary\n\n\u003C3-5 bullet TL;DR derived from README + repo metadata>\n\n## Referenced repos (1-hop)\n\n\u003Cfor each ref with confidence ≥0.6, sorted by verdict severity integrate > take-parts > inspire > skip>\n\n### \u003Cref-full-name> — \u003Cverdict>\n\n- URL: https:\u002F\u002Fgithub.com\u002F\u003Cref>\n- Language: \u003Clang> | Stars: \u003CN> | License: \u003CSPDX>\n- **Rationale:** \u003C1-3 sentences>\n\u003Cif verdict == take-parts>\n- **What to lift:** \u003Cfile\u002Fpath\u002Farea to inspect first>\n\u003C\u002Fif>\n\n## Security scan (advisory)\n\n\u003CPhase 1.6: the scan `summary` line, then one bullet per finding. Always present on a github run.>\n\n## Skill scan (advisory)\n\n\u003COnly present when --skill-scan was passed (Phase 3.7): one line per scanned skill, as the phase reference specifies. If --skill-scan was NOT passed, omit this section entirely.>\n\n## Research enrichment\n\n\u003COnly present when --research was passed (Phase 3.6). 4–8 bullets synthesizing\nthe WebSearch findings about the SOURCE repo — maturity\u002Fadoption, notable\nalternatives, known limitations, community sentiment. Each external claim ends\nwith a recency marker + source URL inline, e.g.:\n`- Actively maintained; v2.3 released as of Mar 2026 ([release notes](https:\u002F\u002F…)).`\nOn dry-run, render `_(dry-run — research not executed)_`. On web failure,\nrender `_(research enrichment requested but web search returned no usable\nresults)_`. If --research was NOT passed, omit this section entirely.>\n\n## Reusable components\n\n\u003COnly present when --deep was passed. Compose from the source-repo scan\nJSON. For each component in `components[]`, one bullet:\n`- **\u003Cname>** (\u003Ctype>) — \u003Cdescription>. _Reuse:_ \u003C1-2 line how-to-rephrase\nfor himmel\u002Fluna — YOUR judgment from the description + type>.`\nThen, for each integrate\u002Ftake-parts ref that was scanned, one summary line:\n`- _Ref \u003Co>\u002F\u003Cr>: N components inventoried → see [[30-Resources\u002FComponents\u002F]]._`\nIf --deep was NOT passed, omit this section entirely.>\n\n## Weak refs (low-confidence, appendix)\n\n\u003Cfor each ref with confidence \u003C0.6 — same structure, condensed>\n\n## Unclassified (transient — re-run to retry)\n\n\u003Cfor each ref with `verdict: api_failure` — list URL + verdict_reason. surfaces rate-limited \u002F network-failed refs distinctly from § Skip so operator knows what to retry. Empty section reads `_(none — all refs classified)_`.>\n\n## Action items\n\n\u003Cextracted from README headings like \"TODO\", \"Roadmap\", \"Known issues\" — convert each to `- [ ] \u003Citem>`. If none found: `_(none — repo README has no TODO\u002FRoadmap section)_`>\n\n## Open questions\n\n\u003Cgaps the chain-follow could not resolve. Always include: \"Verify integrate-verdict refs by reading their root README before taking a dep on any of them.\">\n\n## Audit log\n\n- Refs scanned: \u003CN>\n- Refs above confidence floor (0.6): \u003CK>\n- Refs marked api_failure: \u003CF> (if >0: rc=4; re-run to retry classification)\n- Tail-skipped: \u003CT> (set if more refs existed than --limit)\n- API calls: \u003CN+2> (+1 over the prior count for the `upstream_commit` fetch)\n- README size: \u003Cbytes> (warn if >512 KB; --limit caps API-call count, not text size)\n- Trust tier: \u003Ctier> (\u003Creason>)\n- Safety flag: \u003Cterm-or-none>   (non-blank value reached rc=6 unless --allow-unsafe; document the term in this row regardless)\n```\n\n### Revalidation invariant\n\n**Read on demand:** Only when writing or reviewing a `## Re-validation` section; a fresh ingest never does. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Frevalidation-invariant.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n### Semantic layer (schema defined, NOT yet populated)\n\n**Read on demand:** Only when touching the not-yet-populated semantic fields (`concepts`, `solves`, `maps_to`, `value_tier`, `value_reason`); this MVP never writes them. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fsemantic-layer.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Phase 5 — Daily-note backref (deferred to v2)\n\nMVP does NOT touch the daily note. v2 picks up the obsidian-triage Phase 5 dedup-by-backreference contract.\n\n## Phase 6 — Promotion candidate (deferred to v2)\n\nMVP does NOT write `00-Inbox\u002Fpromote-\u003Cslug>.md`. v2 adds this when any verdict ∈ {integrate, take-parts}.\n\n## Dry-run mode\n\nWhen `--dry-run` is passed:\n- Print \"DRY luna-ingest: would write \u003Cfull-path>\"\n- Print the synthesized markdown body to stdout, prefixed with 4 spaces\n- Print per-ref verdict summary table\n- With `--research`: skip the Phase 3.6 web queries (print `DRY luna-ingest:\n  would run research enrichment for \u003Cowner_repo>`); the `## Research\n  enrichment` section renders `_(dry-run — research not executed)_`.\n- rc=0, touch no files in the vault\n\n## Exit codes\n\n**Read on demand:** Before choosing or interpreting an exit code (0 ok, 1 usage, 2 env, 3 target exists, 4 partial, 5 source fetch failed, 6 safety-blocked). Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fexit-codes.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n\n## Out of scope (v2 follow-ups)\n\n**Read on demand:** Only when asked whether an input form or feature is supported. Read `\u003Cplugin>\u002Freferences\u002Fluna-ingest\u002Fout-of-scope.md` (`\u003Cplugin>` is the plugin root, two directories above this skill's directory) in full and follow it exactly as if it were inlined here; do not work from memory of it.\n",{"name":5,"description":7},"main","2a00f5c1203047a27ae79f5626e1ae4d34f03c79","marketplace\u002Fplugins\u002Fobsidian-triage\u002Fskills\u002Fluna-ingest",2,"https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel","https:\u002F\u002Fyotamleo-org.github.io\u002FHimmel\u002F","https:\u002F\u002Fgithub.com\u002Fyotamleo-org","2026-11-15 05:10:30",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add yotamleo-org\u002FHimmel --skill luna-ingest",{"tool":37,"command":38},"bunx","bunx skills add yotamleo-org\u002FHimmel --skill luna-ingest",{"tool":40,"command":41},"pnpm","pnpm dlx skills add yotamleo-org\u002FHimmel --skill luna-ingest","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fmarketplace\u002Fplugins\u002Fobsidian-triage\u002Fskills\u002Fluna-ingest\u002FSKILL.md",true,[],[46,53,60,67,74,81],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":18},"01M75TY3C1HJV3PACB8R2FTAX7","medic","Use inside a Salus medical vault to FILE a medical item into its structured place (skin photo → photo archive + capture note; lesion\u002Fsymptom note; lab\u002FPDF → 00-Sources) and auto-commit, OR to QUERY the vault for a health follow-up (grounded answer + flagged, non-diagnostic morphology read). Triggers on \"file this photo \u002F note \u002F lab\", \"log this into the medical vault\", \"ask the medic\", or any health follow-up while cwd = the medical vault. Organize & surface only — never diagnose (posture-A).","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Ftemplates\u002Fluna-second-brain\u002F_profiles\u002Fsalus\u002F.claude\u002Fskills\u002Fmedic\u002FSKILL.md",[],[],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":18},"01M75TY1AV2E5DECQFSGZ01FQ3","stuck-playbook","fixture skill, invoked in the basic transcript fixture","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fscripts\u002Flanes\u002Fbench\u002Fscorecard\u002Ffixtures\u002Ftool-usage\u002Fskills\u002F.claude\u002Fskills\u002Fstuck-playbook\u002FSKILL.md",[],[],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":18},"01M75TY05TH00N1KX4H0W1AASQ","never-used-skill","fixture skill, never invoked in any fixture transcript","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fscripts\u002Flanes\u002Fbench\u002Fscorecard\u002Ffixtures\u002Ftool-usage\u002Fskills-config\u002Fskills\u002Fnever-used-skill\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":18},"01M75TXV4XKHHV36JKHPXK0QAN","used-skill","fixture skill invoked via a Skill tool_use in the fixture transcript","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fscripts\u002Flanes\u002Fbench\u002Fscorecard\u002Ffixtures\u002Fdead-parts\u002Fbasic-repo\u002F.claude\u002Fskills\u002Fused-skill\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":18},"01M75TXJYNA5BQWC5WWZN4HY95","dead-skill","fixture skill never invoked and never referenced","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fscripts\u002Flanes\u002Fbench\u002Fscorecard\u002Ffixtures\u002Fdead-parts\u002Fbasic-repo\u002F.claude\u002Fskills\u002Fdead-skill\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":18},"01M75TXAB3RCTMWB02GEKDT17Q","jira-transition","Move\u002Ftransition a Jira issue status (\"move PROJ-46 to In Progress\"). Needs key + target status. Not PR status changes.","https:\u002F\u002Fgithub.com\u002Fyotamleo-org\u002FHimmel\u002Fblob\u002Fmain\u002Fplugins\u002Fhimmel-jira\u002Fskills\u002Fjira-transition\u002FSKILL.md",[],[]]