[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6C298Q7GNPXJ1XSAZ9P2CD3":3,"skill-same:01M6C298Q7GNPXJ1XSAZ9P2CD3":69},{"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":22,"publishedAt":43,"linkOnly":44,"markdownContent":45,"markdownBody":46,"frontMatter":47,"refName":48,"commitSha":49,"skillDirectory":50,"forksCount":51,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":54,"install":55,"resources":68},"01M6C298Q7GNPXJ1XSAZ9P2CD3","nextsteps",null,"Situational assessment and roadmap sync after a work block. Default mode reads ONLY beads (`br`) and `~\u002Froadmap` (lean: independent nextsteps markdown doc + beads update + roadmap activity + ~\u002Froadmap learnings log). `--full` preserves the legacy all-source behavior (adds Claude auto-memory writes, mem0 sync, and GitHub Issue creation). Prefers editing existing roadmap docs over creating new files.","jleechanorg\u002Fjleechan-skills","jleechanorg","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F221404966?v=4","01M6C1VJ99XG950X1DWPEBCBGX","01M6C1VJDK5YVDNT3362RT5WSB",6,"MIT","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fnextsteps\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000031","产品经理","PROFESSION","适用职业",[5,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42],"creation","new","activity","log","learnings","br","roadmap","files","only","default","mode","sync","update","markdown","work","beads","auto","assessment","lean","memory","2026-11-10 10:12:30",false,"---\nname: nextsteps\ndescription: \"Situational assessment and roadmap sync after a work block. Default mode reads ONLY beads (`br`) and `~\u002Froadmap` (lean: independent nextsteps markdown doc + beads update + roadmap activity + ~\u002Froadmap learnings log). `--full` preserves the legacy all-source behavior (adds Claude auto-memory writes, mem0 sync, and GitHub Issue creation). Prefers editing existing roadmap docs over creating new files.\"\n---\n\n# \u002Fnextsteps — Situational Assessment & Roadmap Update\n\nSituational assessment and roadmap sync after a work block.\n\n## Timeline, parallel lanes, and milestones (mandatory)\n\nRead and apply `${CLAUDE_HOME:-$HOME\u002F.claude}\u002Fskills\u002Fparallelize-to-ceiling\u002Freferences\u002Ftimeline-milestones.md`\non every invocation. Always include a timeline, maximize useful independent\nlanes within the measured resource ceiling, and report milestones every\n20 minutes with an hourly rollup during active work. Preserve this command's\nplanning, handoff, and execution authorization boundaries.\n\n## Modes (read first — determines which phases run)\n\n`\u002Fnextsteps` has **two mutually exclusive modes**. Pick one based on the\ninvocation; do not silently blend them.\n\n| Mode | Invocation | Sources read | Side effects written |\n|------|------------|--------------|----------------------|\n| **default** (lean) | `\u002Fnextsteps` or `\u002Fnextsteps [brief]` | beads (`br list`\u002F`br show`) + `~\u002Froadmap\u002F` + `roadmap\u002Factivity\u002FYYYY-MM-DD.md` | nextsteps `.md` doc + `br` updates + `roadmap\u002Factivity\u002FYYYY-MM-DD.md` (and README date link if new date) + `~\u002Froadmap\u002Flearnings-YYYY-MM.md` |\n| **`--full`** (legacy all-source) | `\u002Fnextsteps --full` or `\u002Fnextsteps --full [brief]` | everything in default + Claude auto-memory (`~\u002F.claude\u002Fprojects\u002F\u003Ckey>\u002Fmemory\u002F`) + mem0 | everything in default + Claude auto-memory writes + `MEMORY.md` pointers + `mem0_shared_client.py add` + GitHub Issue creation |\n\n**Default mode skips these phases on purpose** (they are owned by `--full`):\n\n- **Phase 4 — Write to Claude auto-memory** (writes `~\u002F.claude\u002Fprojects\u002F...\u002Fmemory\u002F*.md` and `MEMORY.md` pointers).\n- **Phase 5 — Save to mem0** (calls `~\u002F.hermes\u002Fscripts\u002Fmem0_shared_client.py`).\n- **Phase 7b — Create or update GitHub Issues** (calls `gh issue create` for each new bead).\n\n\u003C!-- USER REQUEST (verbatim, preserved per task brief): \"Make \u002Fnextsteps only do beads and ~\u002Froadmap and \u002Fnextsteps --full does everything\" -->\n\nIf a user wants the side-effecting phases in the default run, they must invoke\n`\u002Fnextsteps --full`. Rationale: the lean default keeps `\u002Fnextsteps` focused on\nbead + roadmap state, which is what the user is asking for; the memory\u002Fissue\nmirroring is opt-in and lives behind `--full`.\n\n### How to parse the flag\n\n1. Look at the literal text right after `\u002Fnextsteps` or in `$ARGUMENTS`.\n2. If the first non-whitespace token is exactly `--full`, run in `--full` mode and strip it from the brief.\n3. Otherwise (no `--full` flag on the invocation), run in default mode and treat the rest as the user-provided brief.\n4. `--full` is the only recognized flag. Any other token (`--help`, `-h`, etc.) is treated as part of the brief.\n\nReport the chosen mode on the **first line of the Phase 8 report**, e.g.:\n\n```\nMode: default (beads + ~\u002Froadmap)\n```\n\nor\n\n```\nMode: --full (beads + ~\u002Froadmap + Claude memory + mem0 + GH Issues)\n```\n\n## Fail-closed rule\n\n## Continue authorized work before handoff\n\nBefore writing a handoff or ending the run, reconcile the live request and\ncurrent task state. A status question, acknowledgement, checkpoint, or context\ntransition does not withdraw authorization for the same action, target, and\noutcome. Complete the next safe, reversible, authorized action—including\navailable reviews and evidence checks—before reporting it as a next step.\nRecord a future task as queued only when it is outside the current request or\nrequires a real gate. Keep explicit planning-only requests, user stops, the\nautonomy deadline, and user-only authority boundaries intact. If blocked, name\nthe exact action, current reason it cannot proceed, controlling owner, attempts\nand alternatives, and a concrete resumption trigger; do not label routine\nverification or review as waiting for the user.\n\n## Roadmap publication\n\nFor routine human-readable updates in `~\u002Froadmap`, follow that repository's\n`AGENTS.md` direct-`origin\u002Fmain` workflow, including scoped mechanical\nvalidation and an expected-head check. Do not open a roadmap PR solely to\npublish a `\u002Fnextsteps` artifact; existing roadmap PRs remain governed by that\nrepository's merge policy. This applies only to roadmap documentation: code\nand skill changes keep their repository PR and merge gates, and Beads state\ncontinues through the canonical `br` publisher.\n\nA `\u002Fnextsteps` run is **incomplete** unless it leaves **all** of these artifacts\n**for its chosen mode**:\n\n### Default mode (lean) — required artifacts\n\n1. **Independent summary markdown doc** (see [Nextsteps document](#nextsteps-document-mandatory)) — TOC, executive summary, then full self-contained detail; bead links throughout\n2. Beads updated\u002Fcreated via `br`\n3. `~\u002Froadmap\u002Flearnings-YYYY-MM.md` appended\n4. **`roadmap\u002Factivity\u002FYYYY-MM-DD.md`** appended with session bullet (repo git root — create file if absent). If that date is brand-new, prepend one date link to `roadmap\u002FREADME.md`'s `## Recent activity (by day)` section.\n\nIf the session has no repo checkout, skip item 4 only and note that in the Phase 8 report.\n\n### `--full` mode — required artifacts\n\nIn addition to all default-mode artifacts above, a `--full` run must also\nleave:\n\n5. Claude memory files written with `MEMORY.md` pointers\n6. mem0 entry saved (or `⚠️ mem0 unavailable (skipped)` recorded)\n7. GitHub Issue creation attempted \u002F recorded\n\nIf the session has no repo checkout, skip item 4 only and note that in the Phase 8 report.\n\n### Continuous completion contract\n\nA normal `\u002Fnextsteps` invocation authorizes all relevant stages below. Complete all required stages in the same run before yielding a final response.\n\n- In **default mode**, execute Stage 1 (Phases 1a\u002F1b, 2, 3), skip Stage 2 side-effecting syncs (owned by `--full`), and produce the Phase 8 report immediately.\n- In **`--full` mode**, execute Stage 1, then proceed continuously into Stage 2 (Claude auto-memory, mem0, GitHub Issues) before producing the Phase 8 report.\n- Do not pause for approval, ask whether to proceed, or stop at stage boundaries.\n- Parallelize independent work when slots are available. If parallel capacity is unavailable, continue locally or sequentially; lack of a subagent slot is never a reason to stop.\n- Interim user updates are non-blocking status messages, not approval checkpoints.\n- If mem0, GitHub Issues, or another optional sink is unavailable in `--full` mode, record the exact attempted result, complete every other artifact, and report the unavailable sink in the final checklist.\n- Stop only for a genuine blocker that requires new authority or information and cannot be safely worked around. Routine external failures, disabled integrations, or partial tool availability are not stopping conditions.\n- Phase 8 completes this command's assessment and persistence work; it does not terminate an ongoing caller workflow or cancel already-authorized work. Return to that workflow and continue its next safe action after reporting.\n\n## Doc discovery — prefer update over create\n\nBefore writing the independent summary or touching roadmap files:\n\n1. **Search for existing nextsteps \u002F session \u002F sync docs** (in order):\n   - Repo: `roadmap\u002Fnextsteps*.md`, `roadmap\u002FNEXT-STEPS.md`, `roadmap\u002Fsession-*.md`, `docs\u002Fnextsteps*.md` (if present)\n   - Home: `~\u002Froadmap\u002Fnextsteps-latest.md`, `~\u002Froadmap\u002Fnextsteps-*.md`\n2. **Prefer:** append a new **dated section** to an existing rolling file (e.g. `nextsteps-latest.md` or the newest `nextsteps-YYYY-MM-DD.md` in the same month), or add a subsection under an existing “Work queue” \u002F “Next steps” heading.\n3. **Create new file only when** no suitable file exists. Default **new** path: `~\u002Froadmap\u002Fnextsteps-\u003CYYYY-MM-DD>.md` (or `roadmap\u002Fnextsteps-\u003CYYYY-MM-DD>.md` if the repo standard is to keep session docs in-repo — match sibling files if any).\n4. **Activity log:** append the session bullet to **`roadmap\u002Factivity\u002FYYYY-MM-DD.md`** (create file with `# Activity — YYYY-MM-DD` header if absent). If this is the first entry for that date, prepend a `- [YYYY-MM-DD](activity\u002FYYYY-MM-DD.md)` line to the `## Recent activity (by day)` list in `roadmap\u002FREADME.md`. Do not prepend to README for subsequent entries on the same date — only the per-day file changes. This keeps README conflict-free across concurrent PRs.\n\n## Nextsteps document (mandatory)\n\nThe independent `.md` file is the **handoff artifact**: a reader must be able to execute the queue without opening beads or chat.\n\n**Document order (fixed):**\n\n1. **Title line** — e.g. `# Nextsteps — \u003Ctopic or repo> — \u003CYYYY-MM-DD>`\n2. **Table of contents** — Markdown list linking to **every** following `##` section (Executive summary through Roadmap pointer). Use GitHub-style anchors (lowercase, hyphenated, dedupe if titles repeat). Update whenever headings change.\n3. **Executive summary** — Short, skimmable block (bullets OK): what this block accomplished, what is blocked or at risk, top priorities \u002F sequencing, beads and PRs that matter (**ids with links**). No deep procedural detail — that lives in the sections below.\n4. **Full detail (all following `##` sections)** — Each section is **self-contained** (definitions, file paths, acceptance criteria, dependencies). Do not rely on “see chat” or unstated context. Order: Context → Bead index → Work queue → Timeline and parallel lanes → PR \u002F merge state → Learnings pointer → Roadmap pointer.\n\n**Required `##` sections after Executive summary (skip only if genuinely N\u002FA — one line stating why):**\n\n| Section | Content |\n|--------|---------|\n| **Context** | 2–6 sentences: what block just ended, repo(s), branch\u002FPR if relevant, scope boundaries |\n| **Bead index** | Table: `bd-…` id, title, priority\u002Fstatus if known, **link** — every open bead touched or created this run. Prefer `https:\u002F\u002Fgithub.com\u002F\u003Cowner>\u002F\u003Crepo>\u002Fissues\u002F\u003Cn>` when the bead syncs to GitHub Issues; else `br show \u003Cid>` as fallback. Link the id in every row. |\n| **Work queue** | Numbered tasks; each task **self-contained**: goal, acceptance criteria, files\u002Fareas, dependencies\u002Fblockers, suggested order; **reference beads** inline as linked `[bd-xxx](url)` where applicable |\n| **Timeline and parallel lanes** | Estimated elapsed windows, lane owners and exclusive write scope, dependencies, outputs\u002Fproof, critical path, concurrency ceiling and limiting resource; 20-minute milestones and hourly rollups for future execution, without resuming stopped work |\n| **PR \u002F merge state** | Same session truth as Phase 1b (`PR #n: OPEN \\| MERGED \\| CLOSED`) for any PR whose current status or merge order is asserted or acted on; explicitly dated historical inventories remain snapshots. Include full PR URLs. |\n| **Learnings pointer** | Path\u002Flink to the new `~\u002Froadmap\u002Flearnings-YYYY-MM.md` entry for this date; one-line summary of what was logged |\n| **Roadmap pointer** | Confirm `roadmap\u002Factivity\u002FYYYY-MM-DD.md` appended (and README date link added if new date) |\n\n**Link rules**\n\n- **Beads:** linked in **Bead index**, and again **inline** in Work queue items where a task maps to a bead.\n- **PRs\u002Fissues:** full `https:\u002F\u002Fgithub.com\u002Fowner\u002Frepo\u002Fpull\u002Fn` (or `\u002Fissues\u002Fn`) when known.\n- **Internal:** link from TOC entries to each `##` section anchor.\n\n**Example skeleton**\n\n```markdown\n# Nextsteps — \u003Crepo> — 2026-04-19\n\n## Table of contents\n\n- [Executive summary](#executive-summary)\n- [Context](#context)\n- [Bead index](#bead-index)\n- [Work queue](#work-queue)\n- [Timeline and parallel lanes](#timeline-and-parallel-lanes)\n- [PR \u002F merge state](#pr--merge-state)\n- [Learnings pointer](#learnings-pointer)\n- [Roadmap pointer](#roadmap-pointer)\n\n## Executive summary\n\n- Outcomes: …\n- Risks \u002F blockers: …\n- Next: …\n- Beads: [bd-abc123](https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fissues\u002FNN) (short label)\n\n## Context\n\n…\n\n## Bead index\n\n| Bead | Title | Link |\n|------|-------|------|\n| bd-… | … | [bd-…](https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fissues\u002FNN) |\n\n## Work queue\n\n1. … — tracks [bd-…](https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fissues\u002FNN)\n\n## Timeline and parallel lanes\n\n| Elapsed estimate | Lane \u002F owner | Scope \u002F dependencies | Deliverable \u002F proof |\n|---|---|---|---|\n| … | … | … | … |\n\n- Critical path and measured or provisional concurrency ceiling: …\n- Milestones: +20m, +40m, +60m (hourly rollup), then repeat while active.\n- Execution start: \u003Cstarted now, or the exact unmet gate and resumption trigger>. Do not mark already-authorized work pending resumption; continue it in the current workflow.\n\n## PR \u002F merge state\n\n- https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fpull\u002F123 — OPEN\n\n## Learnings pointer\n\n- `~\u002Froadmap\u002Flearnings-2026-04.md` — section `2026-04-19 — …`\n\n## Roadmap pointer\n\n- Appended `roadmap\u002Factivity\u002FYYYY-MM-DD.md` — Recent activity (per-day file)\n```\n\n## When invoked — Execution Workflow\n\n### Stage 1: Local Updates (Default & `--full` Modes)\n\n#### Phase 1a — Memory Search Context\nReuse relevant context and source reads already gathered in this session. Delegate\nthis phase alongside Phase 1b only when they contain independent unanswered\nquestions worth separate lanes; otherwise perform the remaining bounded reads\nlocally. Do not repeat completed discovery merely to satisfy a phase label.\nIf a discovery lane fails because its transport or quota is unavailable, finish\nits bounded reads locally rather than retrying that unavailable transport.\nRefresh mutable tracker and PR state when reporting current facts. Refresh a PR\nwhen asserting or acting on its current status or merge order; preserve explicitly\ndated historical inventories as snapshots, including when they cite PRs.\n\nFor context still missing:\n1. Search memory files for key terms from the user-provided context after `\u002Fnextsteps`\n2. Check `~\u002Froadmap\u002Fnextsteps-*.md` for most recent session doc (target for append vs new file)\n3. Check `~\u002Froadmap\u002Flearnings-YYYY-MM.md` tail for existing entries\n4. Report: existing bead IDs, open items from prior sessions, path of most recent nextsteps doc\n\n#### Phase 1b — Gather context\nApply Phase 1a's reuse and delegation decision; share this inspection with other\ncommands composed in the same request. Gather only the context still needed:\n- `git log --oneline -10`\n- `br list --status open --limit 0`\n- `ls roadmap\u002F` (and `ls ~\u002Froadmap\u002F` for home docs)\n- Run [Doc discovery](#doc-discovery--prefer-update-over-create) — note target file for the summary doc (existing vs new path)\n- Use any user-provided line after `\u002Fnextsteps` as extra context.\n- **PR truth (same session):** From open beads, roadmap merge stacks, and user notes, collect every distinct GitHub PR number you will reference. For each `n`, resolve the repo (default: `gh repo view --json nameWithOwner -q .nameWithOwner` from the git root; if the work spans another repo, pass that owner\u002Fname explicitly) and run:\n  ```bash\n  gh pr view \u003Cn> --repo \u003Cowner\u002Frepo> --json state,mergedAt,closedAt,headRefName\n  ```\n  Map JSON to a human line for the report and the **Nextsteps document**: **`PR #n: OPEN`** if `state` is `OPEN`; **`PR #n: MERGED`** if `mergedAt` is non-null; else **`PR #n: CLOSED`** (closed without merge). Do not recommend “land PR *n*” without this check in the **same** `\u002Fnextsteps` run.\n\n#### Phase 2 — Assess & Update Local Assets\n- Match recent commits to open beads; close or update status.\n- Note gaps → determine new beads to create.\n- Create or update beads using `br`:\n  ```bash\n  br create \"\u003Ctitle>\" --type task --priority 2\n  ```\n- Append session bullet to **`roadmap\u002Factivity\u002FYYYY-MM-DD.md`** (create with header if new date). Only touch `roadmap\u002FREADME.md` when the date file is brand-new (prepend one date link to `## Recent activity (by day)`).\n- Identify learnings from the session worth persisting, and append them to `~\u002Froadmap\u002Flearnings-\u003CYYYY-MM>.md` (create if absent) using this format:\n  ```markdown\n  ## \u003CYYYY-MM-DD> — \u003Ctitle>\n  - **Type**: feedback|project|reference\n  - **Classification**: 🚨|⚠️|✅|❌\n  - **Summary**: \u003Cone-liner>\n  - **Bead**: \u003Cbd-id or none>\n  - **Files**: \u003Cpaths changed if any>\n  - **Nextsteps doc**: \u003Cpath to independent summary md>\n  ```\n\n#### Phase 3 — Write\u002FUpdate Nextsteps Document & Show to User\n- Apply [Doc discovery](#doc-discovery--prefer-update-over-create).\n- Write or append to the Nextsteps document: **table of contents**, **executive summary**, then **full detail** sections per [Nextsteps document (mandatory)](#nextsteps-document-mandatory).\n- Record the clean table of created\u002Fupdated beads, learnings, roadmap activity changes, and the Nextsteps document path.\n- In **default mode**, skip Stage 2 and proceed directly to Phase 8 report.\n- In **`--full` mode**, continue directly to Stage 2. Do not yield a final response or request confirmation here.\n\n---\n\n### Stage 2: Memory & Tool Sync (`--full` only, Parallel When Available)\n\n**Skipped by default mode.** Run only when the invocation includes `--full`.\n\nRun the remaining tasks immediately in the same invocation. Use parallel subagents when available; otherwise perform the tasks locally or sequentially and finish the stage.\n\n#### Phase 4 — Write to Claude auto-memory (`--full` only)\nCreate a subagent (`TypeName=self` or a custom subagent) to execute the memory write for each learning\u002Ffinding:\n1. Determine type: `feedback` (rules, anti-patterns) | `project` (decisions, state) | `reference` (pointers)\n2. Slug: lowercase, underscored, max 40 chars\n3. Derive memory dir from git root:\n   ```bash\n   git_root=$(git rev-parse --show-toplevel)\n   project_key=\"${git_root\u002F\u002F\\\u002F\u002F-}\"\n   memory_dir=\"$HOME\u002F.claude\u002Fprojects\u002F${project_key}\u002Fmemory\"\n   ```\n4. Write file `${memory_type}_${date}_${slug}.md` with frontmatter:\n   ```markdown\n   ---\n   name: \u003Ctitle>\n   description: \u003Cone-liner>\n   type: feedback|project|reference\n   bead: \u003Cbd-id or none>\n   ---\n\n   \u003Cbody>\n\n   **Why:** \u003Creason>\n\n   **How to apply:** \u003Cwhen\u002Fwhere this kicks in>\n   ```\n5. Append pointer to `MEMORY.md` (create file if missing): `- [Title](filename) — one-liner`\n6. Report back: `✅ Claude auto-memory: {filename}`\n\n#### Phase 5 — Save to mem0 (`--full` only)\nCreate a subagent (or run concurrently) to save to mem0:\n1. Check: skip if `~\u002F.hermes\u002Fscripts\u002Fmem0_shared_client.py` is absent.\n2. Build text: `\"{title}: {one_liner}. {body_1_sentence}\"`\n3. Run:\n   ```bash\n   python3 ~\u002F.hermes\u002Fscripts\u002Fmem0_shared_client.py add \"\u003Ctext>\" \\\n     --user-id $USER \\\n     --no-infer\n   ```\n4. Report back: `✅ mem0 saved` or `⚠️ mem0 unavailable (skipped)`\n\n#### Phase 7b — Create or update GitHub Issues (`--full` only)\nFor each bead created in Stage 1, attempt to create a linked GitHub Issue. Resolve\nthe repository from the current checkout; this publication is optional, so a\nlookup or creation failure blocks only the issue write. Record the exact failure\nand continue the other Phase 7\u002F8 work. Never choose a different repository as a\nfallback.\n\nRun all issue creates in parallel only after the current repository has been\nresolved:\n```bash\nif ! REPO=$(gh repo view --json nameWithOwner -q .nameWithOwner); then\n  echo \"⚠️ GitHub Issue publication unavailable: repository lookup failed; preserve the exact gh error and continue other artifacts. No issue was created.\"\nelif [[ -z \"$REPO\" ]]; then\n  echo \"⚠️ GitHub Issue publication unavailable: repository lookup returned an empty name. Continue other artifacts. No issue was created.\"\nelse\n  if ! gh issue create --repo \"$REPO\" \\\n      --title \"[resilience] \u003Cbead title>\" \\\n      --body \"## Summary\n\u003Cone-paragraph description>\n\n## Acceptance criteria\n- [ ] \u003Ccriterion 1>\n- [ ] \u003Ccriterion 2>\n\n**Bead:** \u003Cbd-id>\" \\\n      --label \"task\" 2>&1; then\n    echo \"⚠️ GitHub Issue publication unavailable: preserve the exact gh error; do not retry in another repository. Continue other artifacts.\"\n  fi\nfi\n```\n- If `gh issue create` returns \"repository has disabled issues\", log `⚠️ GH Issues disabled on \u003Crepo> — bead only` and continue.\n- If repository lookup or issue creation fails for another reason, retain its exact error and leave that optional issue unpublished; never retry in a guessed repository.\n- If it succeeds, capture the issue URL and add it to the bead index in the Nextsteps document.\n- Report back: `✅ GH Issue #N created`, `⚠️ GH Issues disabled`, or `⚠️ GitHub Issue publication unavailable: \u003Cexact reason>`.\n\n---\n\n### Phase 8 — Report completion\n\n**First line:** the chosen mode (see [Modes](#modes-read-first--determines-which-phases-run)).\n\nThen summarize the results and print the final artifact checklist **for the chosen mode**:\n\n**Default mode checklist:**\n\n- `[x]` **Nextsteps independent `.md`** (TOC + executive summary + full detail; bead index + linked beads in queue)\n- `[x]` Beads (`br`) written\n- `[x]` `~\u002Froadmap\u002Flearnings-YYYY-MM.md` updated (includes nextsteps doc path)\n- `[x]` `roadmap\u002Factivity\u002FYYYY-MM-DD.md` (and `roadmap\u002FREADME.md` if new date) updated\n- `[ ]` (intentionally blank — Claude memory + mem0 + GH Issues are owned by `--full`)\n\n**`--full` mode checklist:**\n\n- `[x]` **Nextsteps independent `.md`** (TOC + executive summary + full detail; bead index + linked beads in queue)\n- `[x]` Beads (`br`) written\n- `[x]` Claude memory + `MEMORY.md` pointers written\n- `[x]` `~\u002Froadmap\u002Flearnings-YYYY-MM.md` updated (includes nextsteps doc path)\n- `[x]` `roadmap\u002Factivity\u002FYYYY-MM-DD.md` (and `roadmap\u002FREADME.md` if new date) updated\n- `[x]` mem0 entry saved (or `⚠️ mem0 unavailable (skipped)`)\n- `[x]` GH Issues created (or `⚠️ GH Issues disabled — bead only`)\n\n**Merge-order sanity:** If any recommended action was “merge **A** before **B**” (or “land A then rebase B”), re-assert **A** using the Phase 1 `gh pr view` results from **this** run. If **A** is **MERGED**, do **not** tell the reader to land A; say instead to **rebase B on `main`** (or the appropriate default branch). If **A** is **OPEN**, keep the ordering advice. If **A** is **CLOSED** without merge, drop merge-order advice and flag that the stack needs re-triage.\n","# \u002Fnextsteps — Situational Assessment & Roadmap Update\n\nSituational assessment and roadmap sync after a work block.\n\n## Timeline, parallel lanes, and milestones (mandatory)\n\nRead and apply `${CLAUDE_HOME:-$HOME\u002F.claude}\u002Fskills\u002Fparallelize-to-ceiling\u002Freferences\u002Ftimeline-milestones.md`\non every invocation. Always include a timeline, maximize useful independent\nlanes within the measured resource ceiling, and report milestones every\n20 minutes with an hourly rollup during active work. Preserve this command's\nplanning, handoff, and execution authorization boundaries.\n\n## Modes (read first — determines which phases run)\n\n`\u002Fnextsteps` has **two mutually exclusive modes**. Pick one based on the\ninvocation; do not silently blend them.\n\n| Mode | Invocation | Sources read | Side effects written |\n|------|------------|--------------|----------------------|\n| **default** (lean) | `\u002Fnextsteps` or `\u002Fnextsteps [brief]` | beads (`br list`\u002F`br show`) + `~\u002Froadmap\u002F` + `roadmap\u002Factivity\u002FYYYY-MM-DD.md` | nextsteps `.md` doc + `br` updates + `roadmap\u002Factivity\u002FYYYY-MM-DD.md` (and README date link if new date) + `~\u002Froadmap\u002Flearnings-YYYY-MM.md` |\n| **`--full`** (legacy all-source) | `\u002Fnextsteps --full` or `\u002Fnextsteps --full [brief]` | everything in default + Claude auto-memory (`~\u002F.claude\u002Fprojects\u002F\u003Ckey>\u002Fmemory\u002F`) + mem0 | everything in default + Claude auto-memory writes + `MEMORY.md` pointers + `mem0_shared_client.py add` + GitHub Issue creation |\n\n**Default mode skips these phases on purpose** (they are owned by `--full`):\n\n- **Phase 4 — Write to Claude auto-memory** (writes `~\u002F.claude\u002Fprojects\u002F...\u002Fmemory\u002F*.md` and `MEMORY.md` pointers).\n- **Phase 5 — Save to mem0** (calls `~\u002F.hermes\u002Fscripts\u002Fmem0_shared_client.py`).\n- **Phase 7b — Create or update GitHub Issues** (calls `gh issue create` for each new bead).\n\n\u003C!-- USER REQUEST (verbatim, preserved per task brief): \"Make \u002Fnextsteps only do beads and ~\u002Froadmap and \u002Fnextsteps --full does everything\" -->\n\nIf a user wants the side-effecting phases in the default run, they must invoke\n`\u002Fnextsteps --full`. Rationale: the lean default keeps `\u002Fnextsteps` focused on\nbead + roadmap state, which is what the user is asking for; the memory\u002Fissue\nmirroring is opt-in and lives behind `--full`.\n\n### How to parse the flag\n\n1. Look at the literal text right after `\u002Fnextsteps` or in `$ARGUMENTS`.\n2. If the first non-whitespace token is exactly `--full`, run in `--full` mode and strip it from the brief.\n3. Otherwise (no `--full` flag on the invocation), run in default mode and treat the rest as the user-provided brief.\n4. `--full` is the only recognized flag. Any other token (`--help`, `-h`, etc.) is treated as part of the brief.\n\nReport the chosen mode on the **first line of the Phase 8 report**, e.g.:\n\n```\nMode: default (beads + ~\u002Froadmap)\n```\n\nor\n\n```\nMode: --full (beads + ~\u002Froadmap + Claude memory + mem0 + GH Issues)\n```\n\n## Fail-closed rule\n\n## Continue authorized work before handoff\n\nBefore writing a handoff or ending the run, reconcile the live request and\ncurrent task state. A status question, acknowledgement, checkpoint, or context\ntransition does not withdraw authorization for the same action, target, and\noutcome. Complete the next safe, reversible, authorized action—including\navailable reviews and evidence checks—before reporting it as a next step.\nRecord a future task as queued only when it is outside the current request or\nrequires a real gate. Keep explicit planning-only requests, user stops, the\nautonomy deadline, and user-only authority boundaries intact. If blocked, name\nthe exact action, current reason it cannot proceed, controlling owner, attempts\nand alternatives, and a concrete resumption trigger; do not label routine\nverification or review as waiting for the user.\n\n## Roadmap publication\n\nFor routine human-readable updates in `~\u002Froadmap`, follow that repository's\n`AGENTS.md` direct-`origin\u002Fmain` workflow, including scoped mechanical\nvalidation and an expected-head check. Do not open a roadmap PR solely to\npublish a `\u002Fnextsteps` artifact; existing roadmap PRs remain governed by that\nrepository's merge policy. This applies only to roadmap documentation: code\nand skill changes keep their repository PR and merge gates, and Beads state\ncontinues through the canonical `br` publisher.\n\nA `\u002Fnextsteps` run is **incomplete** unless it leaves **all** of these artifacts\n**for its chosen mode**:\n\n### Default mode (lean) — required artifacts\n\n1. **Independent summary markdown doc** (see [Nextsteps document](#nextsteps-document-mandatory)) — TOC, executive summary, then full self-contained detail; bead links throughout\n2. Beads updated\u002Fcreated via `br`\n3. `~\u002Froadmap\u002Flearnings-YYYY-MM.md` appended\n4. **`roadmap\u002Factivity\u002FYYYY-MM-DD.md`** appended with session bullet (repo git root — create file if absent). If that date is brand-new, prepend one date link to `roadmap\u002FREADME.md`'s `## Recent activity (by day)` section.\n\nIf the session has no repo checkout, skip item 4 only and note that in the Phase 8 report.\n\n### `--full` mode — required artifacts\n\nIn addition to all default-mode artifacts above, a `--full` run must also\nleave:\n\n5. Claude memory files written with `MEMORY.md` pointers\n6. mem0 entry saved (or `⚠️ mem0 unavailable (skipped)` recorded)\n7. GitHub Issue creation attempted \u002F recorded\n\nIf the session has no repo checkout, skip item 4 only and note that in the Phase 8 report.\n\n### Continuous completion contract\n\nA normal `\u002Fnextsteps` invocation authorizes all relevant stages below. Complete all required stages in the same run before yielding a final response.\n\n- In **default mode**, execute Stage 1 (Phases 1a\u002F1b, 2, 3), skip Stage 2 side-effecting syncs (owned by `--full`), and produce the Phase 8 report immediately.\n- In **`--full` mode**, execute Stage 1, then proceed continuously into Stage 2 (Claude auto-memory, mem0, GitHub Issues) before producing the Phase 8 report.\n- Do not pause for approval, ask whether to proceed, or stop at stage boundaries.\n- Parallelize independent work when slots are available. If parallel capacity is unavailable, continue locally or sequentially; lack of a subagent slot is never a reason to stop.\n- Interim user updates are non-blocking status messages, not approval checkpoints.\n- If mem0, GitHub Issues, or another optional sink is unavailable in `--full` mode, record the exact attempted result, complete every other artifact, and report the unavailable sink in the final checklist.\n- Stop only for a genuine blocker that requires new authority or information and cannot be safely worked around. Routine external failures, disabled integrations, or partial tool availability are not stopping conditions.\n- Phase 8 completes this command's assessment and persistence work; it does not terminate an ongoing caller workflow or cancel already-authorized work. Return to that workflow and continue its next safe action after reporting.\n\n## Doc discovery — prefer update over create\n\nBefore writing the independent summary or touching roadmap files:\n\n1. **Search for existing nextsteps \u002F session \u002F sync docs** (in order):\n   - Repo: `roadmap\u002Fnextsteps*.md`, `roadmap\u002FNEXT-STEPS.md`, `roadmap\u002Fsession-*.md`, `docs\u002Fnextsteps*.md` (if present)\n   - Home: `~\u002Froadmap\u002Fnextsteps-latest.md`, `~\u002Froadmap\u002Fnextsteps-*.md`\n2. **Prefer:** append a new **dated section** to an existing rolling file (e.g. `nextsteps-latest.md` or the newest `nextsteps-YYYY-MM-DD.md` in the same month), or add a subsection under an existing “Work queue” \u002F “Next steps” heading.\n3. **Create new file only when** no suitable file exists. Default **new** path: `~\u002Froadmap\u002Fnextsteps-\u003CYYYY-MM-DD>.md` (or `roadmap\u002Fnextsteps-\u003CYYYY-MM-DD>.md` if the repo standard is to keep session docs in-repo — match sibling files if any).\n4. **Activity log:** append the session bullet to **`roadmap\u002Factivity\u002FYYYY-MM-DD.md`** (create file with `# Activity — YYYY-MM-DD` header if absent). If this is the first entry for that date, prepend a `- [YYYY-MM-DD](activity\u002FYYYY-MM-DD.md)` line to the `## Recent activity (by day)` list in `roadmap\u002FREADME.md`. Do not prepend to README for subsequent entries on the same date — only the per-day file changes. This keeps README conflict-free across concurrent PRs.\n\n## Nextsteps document (mandatory)\n\nThe independent `.md` file is the **handoff artifact**: a reader must be able to execute the queue without opening beads or chat.\n\n**Document order (fixed):**\n\n1. **Title line** — e.g. `# Nextsteps — \u003Ctopic or repo> — \u003CYYYY-MM-DD>`\n2. **Table of contents** — Markdown list linking to **every** following `##` section (Executive summary through Roadmap pointer). Use GitHub-style anchors (lowercase, hyphenated, dedupe if titles repeat). Update whenever headings change.\n3. **Executive summary** — Short, skimmable block (bullets OK): what this block accomplished, what is blocked or at risk, top priorities \u002F sequencing, beads and PRs that matter (**ids with links**). No deep procedural detail — that lives in the sections below.\n4. **Full detail (all following `##` sections)** — Each section is **self-contained** (definitions, file paths, acceptance criteria, dependencies). Do not rely on “see chat” or unstated context. Order: Context → Bead index → Work queue → Timeline and parallel lanes → PR \u002F merge state → Learnings pointer → Roadmap pointer.\n\n**Required `##` sections after Executive summary (skip only if genuinely N\u002FA — one line stating why):**\n\n| Section | Content |\n|--------|---------|\n| **Context** | 2–6 sentences: what block just ended, repo(s), branch\u002FPR if relevant, scope boundaries |\n| **Bead index** | Table: `bd-…` id, title, priority\u002Fstatus if known, **link** — every open bead touched or created this run. Prefer `https:\u002F\u002Fgithub.com\u002F\u003Cowner>\u002F\u003Crepo>\u002Fissues\u002F\u003Cn>` when the bead syncs to GitHub Issues; else `br show \u003Cid>` as fallback. Link the id in every row. |\n| **Work queue** | Numbered tasks; each task **self-contained**: goal, acceptance criteria, files\u002Fareas, dependencies\u002Fblockers, suggested order; **reference beads** inline as linked `[bd-xxx](url)` where applicable |\n| **Timeline and parallel lanes** | Estimated elapsed windows, lane owners and exclusive write scope, dependencies, outputs\u002Fproof, critical path, concurrency ceiling and limiting resource; 20-minute milestones and hourly rollups for future execution, without resuming stopped work |\n| **PR \u002F merge state** | Same session truth as Phase 1b (`PR #n: OPEN \\| MERGED \\| CLOSED`) for any PR whose current status or merge order is asserted or acted on; explicitly dated historical inventories remain snapshots. Include full PR URLs. |\n| **Learnings pointer** | Path\u002Flink to the new `~\u002Froadmap\u002Flearnings-YYYY-MM.md` entry for this date; one-line summary of what was logged |\n| **Roadmap pointer** | Confirm `roadmap\u002Factivity\u002FYYYY-MM-DD.md` appended (and README date link added if new date) |\n\n**Link rules**\n\n- **Beads:** linked in **Bead index**, and again **inline** in Work queue items where a task maps to a bead.\n- **PRs\u002Fissues:** full `https:\u002F\u002Fgithub.com\u002Fowner\u002Frepo\u002Fpull\u002Fn` (or `\u002Fissues\u002Fn`) when known.\n- **Internal:** link from TOC entries to each `##` section anchor.\n\n**Example skeleton**\n\n```markdown\n# Nextsteps — \u003Crepo> — 2026-04-19\n\n## Table of contents\n\n- [Executive summary](#executive-summary)\n- [Context](#context)\n- [Bead index](#bead-index)\n- [Work queue](#work-queue)\n- [Timeline and parallel lanes](#timeline-and-parallel-lanes)\n- [PR \u002F merge state](#pr--merge-state)\n- [Learnings pointer](#learnings-pointer)\n- [Roadmap pointer](#roadmap-pointer)\n\n## Executive summary\n\n- Outcomes: …\n- Risks \u002F blockers: …\n- Next: …\n- Beads: [bd-abc123](https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fissues\u002FNN) (short label)\n\n## Context\n\n…\n\n## Bead index\n\n| Bead | Title | Link |\n|------|-------|------|\n| bd-… | … | [bd-…](https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fissues\u002FNN) |\n\n## Work queue\n\n1. … — tracks [bd-…](https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fissues\u002FNN)\n\n## Timeline and parallel lanes\n\n| Elapsed estimate | Lane \u002F owner | Scope \u002F dependencies | Deliverable \u002F proof |\n|---|---|---|---|\n| … | … | … | … |\n\n- Critical path and measured or provisional concurrency ceiling: …\n- Milestones: +20m, +40m, +60m (hourly rollup), then repeat while active.\n- Execution start: \u003Cstarted now, or the exact unmet gate and resumption trigger>. Do not mark already-authorized work pending resumption; continue it in the current workflow.\n\n## PR \u002F merge state\n\n- https:\u002F\u002Fgithub.com\u002Forg\u002Frepo\u002Fpull\u002F123 — OPEN\n\n## Learnings pointer\n\n- `~\u002Froadmap\u002Flearnings-2026-04.md` — section `2026-04-19 — …`\n\n## Roadmap pointer\n\n- Appended `roadmap\u002Factivity\u002FYYYY-MM-DD.md` — Recent activity (per-day file)\n```\n\n## When invoked — Execution Workflow\n\n### Stage 1: Local Updates (Default & `--full` Modes)\n\n#### Phase 1a — Memory Search Context\nReuse relevant context and source reads already gathered in this session. Delegate\nthis phase alongside Phase 1b only when they contain independent unanswered\nquestions worth separate lanes; otherwise perform the remaining bounded reads\nlocally. Do not repeat completed discovery merely to satisfy a phase label.\nIf a discovery lane fails because its transport or quota is unavailable, finish\nits bounded reads locally rather than retrying that unavailable transport.\nRefresh mutable tracker and PR state when reporting current facts. Refresh a PR\nwhen asserting or acting on its current status or merge order; preserve explicitly\ndated historical inventories as snapshots, including when they cite PRs.\n\nFor context still missing:\n1. Search memory files for key terms from the user-provided context after `\u002Fnextsteps`\n2. Check `~\u002Froadmap\u002Fnextsteps-*.md` for most recent session doc (target for append vs new file)\n3. Check `~\u002Froadmap\u002Flearnings-YYYY-MM.md` tail for existing entries\n4. Report: existing bead IDs, open items from prior sessions, path of most recent nextsteps doc\n\n#### Phase 1b — Gather context\nApply Phase 1a's reuse and delegation decision; share this inspection with other\ncommands composed in the same request. Gather only the context still needed:\n- `git log --oneline -10`\n- `br list --status open --limit 0`\n- `ls roadmap\u002F` (and `ls ~\u002Froadmap\u002F` for home docs)\n- Run [Doc discovery](#doc-discovery--prefer-update-over-create) — note target file for the summary doc (existing vs new path)\n- Use any user-provided line after `\u002Fnextsteps` as extra context.\n- **PR truth (same session):** From open beads, roadmap merge stacks, and user notes, collect every distinct GitHub PR number you will reference. For each `n`, resolve the repo (default: `gh repo view --json nameWithOwner -q .nameWithOwner` from the git root; if the work spans another repo, pass that owner\u002Fname explicitly) and run:\n  ```bash\n  gh pr view \u003Cn> --repo \u003Cowner\u002Frepo> --json state,mergedAt,closedAt,headRefName\n  ```\n  Map JSON to a human line for the report and the **Nextsteps document**: **`PR #n: OPEN`** if `state` is `OPEN`; **`PR #n: MERGED`** if `mergedAt` is non-null; else **`PR #n: CLOSED`** (closed without merge). Do not recommend “land PR *n*” without this check in the **same** `\u002Fnextsteps` run.\n\n#### Phase 2 — Assess & Update Local Assets\n- Match recent commits to open beads; close or update status.\n- Note gaps → determine new beads to create.\n- Create or update beads using `br`:\n  ```bash\n  br create \"\u003Ctitle>\" --type task --priority 2\n  ```\n- Append session bullet to **`roadmap\u002Factivity\u002FYYYY-MM-DD.md`** (create with header if new date). Only touch `roadmap\u002FREADME.md` when the date file is brand-new (prepend one date link to `## Recent activity (by day)`).\n- Identify learnings from the session worth persisting, and append them to `~\u002Froadmap\u002Flearnings-\u003CYYYY-MM>.md` (create if absent) using this format:\n  ```markdown\n  ## \u003CYYYY-MM-DD> — \u003Ctitle>\n  - **Type**: feedback|project|reference\n  - **Classification**: 🚨|⚠️|✅|❌\n  - **Summary**: \u003Cone-liner>\n  - **Bead**: \u003Cbd-id or none>\n  - **Files**: \u003Cpaths changed if any>\n  - **Nextsteps doc**: \u003Cpath to independent summary md>\n  ```\n\n#### Phase 3 — Write\u002FUpdate Nextsteps Document & Show to User\n- Apply [Doc discovery](#doc-discovery--prefer-update-over-create).\n- Write or append to the Nextsteps document: **table of contents**, **executive summary**, then **full detail** sections per [Nextsteps document (mandatory)](#nextsteps-document-mandatory).\n- Record the clean table of created\u002Fupdated beads, learnings, roadmap activity changes, and the Nextsteps document path.\n- In **default mode**, skip Stage 2 and proceed directly to Phase 8 report.\n- In **`--full` mode**, continue directly to Stage 2. Do not yield a final response or request confirmation here.\n\n---\n\n### Stage 2: Memory & Tool Sync (`--full` only, Parallel When Available)\n\n**Skipped by default mode.** Run only when the invocation includes `--full`.\n\nRun the remaining tasks immediately in the same invocation. Use parallel subagents when available; otherwise perform the tasks locally or sequentially and finish the stage.\n\n#### Phase 4 — Write to Claude auto-memory (`--full` only)\nCreate a subagent (`TypeName=self` or a custom subagent) to execute the memory write for each learning\u002Ffinding:\n1. Determine type: `feedback` (rules, anti-patterns) | `project` (decisions, state) | `reference` (pointers)\n2. Slug: lowercase, underscored, max 40 chars\n3. Derive memory dir from git root:\n   ```bash\n   git_root=$(git rev-parse --show-toplevel)\n   project_key=\"${git_root\u002F\u002F\\\u002F\u002F-}\"\n   memory_dir=\"$HOME\u002F.claude\u002Fprojects\u002F${project_key}\u002Fmemory\"\n   ```\n4. Write file `${memory_type}_${date}_${slug}.md` with frontmatter:\n   ```markdown\n   ---\n   name: \u003Ctitle>\n   description: \u003Cone-liner>\n   type: feedback|project|reference\n   bead: \u003Cbd-id or none>\n   ---\n\n   \u003Cbody>\n\n   **Why:** \u003Creason>\n\n   **How to apply:** \u003Cwhen\u002Fwhere this kicks in>\n   ```\n5. Append pointer to `MEMORY.md` (create file if missing): `- [Title](filename) — one-liner`\n6. Report back: `✅ Claude auto-memory: {filename}`\n\n#### Phase 5 — Save to mem0 (`--full` only)\nCreate a subagent (or run concurrently) to save to mem0:\n1. Check: skip if `~\u002F.hermes\u002Fscripts\u002Fmem0_shared_client.py` is absent.\n2. Build text: `\"{title}: {one_liner}. {body_1_sentence}\"`\n3. Run:\n   ```bash\n   python3 ~\u002F.hermes\u002Fscripts\u002Fmem0_shared_client.py add \"\u003Ctext>\" \\\n     --user-id $USER \\\n     --no-infer\n   ```\n4. Report back: `✅ mem0 saved` or `⚠️ mem0 unavailable (skipped)`\n\n#### Phase 7b — Create or update GitHub Issues (`--full` only)\nFor each bead created in Stage 1, attempt to create a linked GitHub Issue. Resolve\nthe repository from the current checkout; this publication is optional, so a\nlookup or creation failure blocks only the issue write. Record the exact failure\nand continue the other Phase 7\u002F8 work. Never choose a different repository as a\nfallback.\n\nRun all issue creates in parallel only after the current repository has been\nresolved:\n```bash\nif ! REPO=$(gh repo view --json nameWithOwner -q .nameWithOwner); then\n  echo \"⚠️ GitHub Issue publication unavailable: repository lookup failed; preserve the exact gh error and continue other artifacts. No issue was created.\"\nelif [[ -z \"$REPO\" ]]; then\n  echo \"⚠️ GitHub Issue publication unavailable: repository lookup returned an empty name. Continue other artifacts. No issue was created.\"\nelse\n  if ! gh issue create --repo \"$REPO\" \\\n      --title \"[resilience] \u003Cbead title>\" \\\n      --body \"## Summary\n\u003Cone-paragraph description>\n\n## Acceptance criteria\n- [ ] \u003Ccriterion 1>\n- [ ] \u003Ccriterion 2>\n\n**Bead:** \u003Cbd-id>\" \\\n      --label \"task\" 2>&1; then\n    echo \"⚠️ GitHub Issue publication unavailable: preserve the exact gh error; do not retry in another repository. Continue other artifacts.\"\n  fi\nfi\n```\n- If `gh issue create` returns \"repository has disabled issues\", log `⚠️ GH Issues disabled on \u003Crepo> — bead only` and continue.\n- If repository lookup or issue creation fails for another reason, retain its exact error and leave that optional issue unpublished; never retry in a guessed repository.\n- If it succeeds, capture the issue URL and add it to the bead index in the Nextsteps document.\n- Report back: `✅ GH Issue #N created`, `⚠️ GH Issues disabled`, or `⚠️ GitHub Issue publication unavailable: \u003Cexact reason>`.\n\n---\n\n### Phase 8 — Report completion\n\n**First line:** the chosen mode (see [Modes](#modes-read-first--determines-which-phases-run)).\n\nThen summarize the results and print the final artifact checklist **for the chosen mode**:\n\n**Default mode checklist:**\n\n- `[x]` **Nextsteps independent `.md`** (TOC + executive summary + full detail; bead index + linked beads in queue)\n- `[x]` Beads (`br`) written\n- `[x]` `~\u002Froadmap\u002Flearnings-YYYY-MM.md` updated (includes nextsteps doc path)\n- `[x]` `roadmap\u002Factivity\u002FYYYY-MM-DD.md` (and `roadmap\u002FREADME.md` if new date) updated\n- `[ ]` (intentionally blank — Claude memory + mem0 + GH Issues are owned by `--full`)\n\n**`--full` mode checklist:**\n\n- `[x]` **Nextsteps independent `.md`** (TOC + executive summary + full detail; bead index + linked beads in queue)\n- `[x]` Beads (`br`) written\n- `[x]` Claude memory + `MEMORY.md` pointers written\n- `[x]` `~\u002Froadmap\u002Flearnings-YYYY-MM.md` updated (includes nextsteps doc path)\n- `[x]` `roadmap\u002Factivity\u002FYYYY-MM-DD.md` (and `roadmap\u002FREADME.md` if new date) updated\n- `[x]` mem0 entry saved (or `⚠️ mem0 unavailable (skipped)`)\n- `[x]` GH Issues created (or `⚠️ GH Issues disabled — bead only`)\n\n**Merge-order sanity:** If any recommended action was “merge **A** before **B**” (or “land A then rebase B”), re-assert **A** using the Phase 1 `gh pr view` results from **this** run. If **A** is **MERGED**, do **not** tell the reader to land A; say instead to **rebase B on `main`** (or the appropriate default branch). If **A** is **OPEN**, keep the ordering advice. If **A** is **CLOSED** without merge, drop merge-order advice and flag that the stack needs re-triage.\n",{"name":5,"description":7},"main","2d92c349286cf4e1cfed8cdec1773e71f6c601d2",".claude\u002Fskills\u002Fnextsteps",0,"https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills","https:\u002F\u002Fgithub.com\u002Fjleechanorg","2026-11-14 03:18:13",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add jleechanorg\u002Fjleechan-skills --skill nextsteps",{"tool":61,"command":62},"bunx","bunx skills add jleechanorg\u002Fjleechan-skills --skill nextsteps",{"tool":64,"command":65},"pnpm","pnpm dlx skills add jleechanorg\u002Fjleechan-skills --skill nextsteps","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fnextsteps\u002FSKILL.md",true,[],[70,88,108,122,136,150],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":43},"01M716Y5MTJ6GYJZ2ATXCJK8G4","instinct","Use when the user invokes \u002Finstinct or \u002Fin, or asks to \"message Instinct\", \"ask Instinct\", \"tell Instinct\", \"check Instinct\", or read replies from the configured peer assistant. iMessage first, WhatsApp Web as fallback.","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Finstinct\u002FSKILL.md",[],[72,77,78,79,80,81,82,83,84,85,86,87],"user","when","replies","check","use","identity","whatsapp","gmail","read","message","ask",{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":97,"publishedAt":43},"01M716Y4HF7FVVMJSA2CCVN5VF","hr","Compatibility alias for history-resume; produce the same fresh-state, provenance-aware resume briefing.","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fhr\u002FSKILL.md",[94],{"categoryNo":95,"categoryName":96,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000038","人力资源",[90,98,99,100,101,102,103,104,105,106,107],"resume","compatibility","provenance","for","history","produce","alias","briefing","state","fresh",{"skillNo":109,"name":110,"nameZh":6,"description":111,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":115,"publishedAt":43},"01M716Y1FFNFKFJYYKF8HQA6T8","history-resume","Resume briefing for the current repository branch or worktree using fresh state and admitted history sources.","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fhistory-resume\u002FSKILL.md",[114],{"categoryNo":95,"categoryName":96,"taxonomyType":20,"taxonomyTypeDesc":21},[102,98,116,117,77,118,105,119,101,120,121,106,107],"history resume","branch","analysis","repository","worktree","sources",{"skillNo":123,"name":124,"nameZh":6,"description":125,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":126,"categories":127,"keywords":128,"publishedAt":43},"01M716XWXJ8EPHK4RH9T14TR85","grok","Alias for the canonical grokbot browser conversation workflow; \u002Fgrok talks to the Grok bot the same way \u002Fdot talks to Dot.","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fgrok\u002FSKILL.md",[],[124,129,130,131,101,132,104,133,134,135],"dot","bot","workflow","canonical","conversation","browser","grokbot",{"skillNo":137,"name":138,"nameZh":6,"description":139,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":140,"categories":141,"keywords":142,"publishedAt":43},"01M70N9Q07DTE0EW5KFXMMMSJ0","codexa-bench","Compare `codexa` and normal `codex` on a fair split of comparable work by default; use `--strict` for duplicate runs of the exact same task.","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fcodexa-bench\u002FSKILL.md",[],[143,37,144,81,32,101,145,146,147,148,149],"codexa bench","split","duplicate","task","compare","normal","codex",{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":154,"licenseFileUrl":155,"sourceHtmlUrl":156,"categories":157,"keywords":169,"publishedAt":43},"01M6C24S2X60SP8BBFGAEKPEGW","frontend-design","Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, React components, HTML\u002FCSS layouts, or when styling\u002Fbeautifying any web UI). Generates creative, polished code and UI design that avoids generic AI aesthetics.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Ffrontend-design\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fjleechanorg\u002Fjleechan-skills\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Ffrontend-design\u002FSKILL.md",[158,163,166],{"categoryNo":159,"categoryName":160,"taxonomyType":161,"taxonomyTypeDesc":162},"01JBSEEDCAT000000000000009","前端开发","SCENARIO","使用场景",{"categoryNo":164,"categoryName":165,"taxonomyType":161,"taxonomyTypeDesc":162},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":167,"categoryName":168,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000032","设计师",[170,171,172,81,173,174,175,77,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,191,192,193,78,194],"frontend","design","frontend design","generic","ui design","react","html","components","code","artifacts","ui","aesthetics","Production","web components","landing","applications","quality","web","css","create","examples","with","react components","web ui","pages"]