[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72XVVN2P71ZMCB2Z004VFSX":3,"skill-same:01M72XVVN2P71ZMCB2Z004VFSX":52},{"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":31,"commitSha":32,"skillDirectory":33,"forksCount":34,"repositoryHtmlUrl":35,"repositoryHomepageUrl":36,"ownerHtmlUrl":37,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":38,"install":39,"resources":51},"01M72XVVN2P71ZMCB2Z004VFSX","next",null,"Use when user asks \"what's next\" or \"next step\". Prints next slash command from current state via a sonnet subagent. Flags: `--clear`, `--list`.","grimoire-rs\u002Fgrimoire","grimoire-rs","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F298895348?v=4","01M72XK7XKYKMET2MMHCAQ3X6C","01M72XK7Y647QCPS3Z02ZYTJTT",15,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fnext\u002FSKILL.md",[],[],"2026-11-10 22:39:42",false,"---\nname: next\ndescription: 'Use when user asks \"what''s next\" or \"next step\". Prints next slash command from current state via a sonnet subagent. Flags: `--clear`, `--list`.'\nuser-invocable: true\ndisable-model-invocation: true\nargument-hint: \"[--clear | --list]\"\ntriggers:\n  - \"what's next\"\n  - \"what is next\"\n  - \"next step\"\n  - \"next command\"\n  - \"what now\"\n---\n\n# \u002Fnext — Suggest the Next Slash Command\n\nLook at current state (branch, working tree, plans, recent commits, open PR). Print exactly **one** slash command user copy-paste to continue. State inspection delegated to **sonnet subagent** so output stay compact, main session context not flooded with `git log` \u002F `gh pr view` \u002F plan-file output.\n\n**Primarily read-only.** Becomes a one-shot **state-fixer** when no `## Status` block found and no `.claude\u002Fstate\u002Fcurrent_plan.md` pointer — prompts user to confirm inferred state, then writes `current_plan.md` + injects Status block into plan so future invocations land in the fast path. Never edits code, never commits, never pushes.\n\n## Flags\n\n- `--clear` — tell subagent make suggested command **fully self-contained** (absolute paths, no \"the work we just did\" refs), prepend `\u002Fclear` so future session with no memory run cleanly.\n- `--list` — subagent return top 3 candidates with one-line rationale each.\n- (no flag) — one self-contained command + one-line \"why\" + one-line \"what was checked\".\n\n## Workflow\n\n### 1. Parse flags\n\nStrip `--clear` and `--list` from argument string. Leftover = free-text intent (rare; e.g. `\u002Fnext --clear after fixing the bug`).\n\n### 2. Delegate to sonnet subagent\n\nSpawn **one** subagent (`Agent` tool, `subagent_type: general-purpose`, `model: sonnet`) with prompt below. Do NOT run inspection commands yourself — point of delegation = keep main context clean.\n\n**Subagent prompt template** (fill `{flags}` and `{intent}` placeholders):\n\n```\nYou are inspecting Grimoire project state to suggest exactly ONE next slash command for the user to type. You are read-only — never edit, commit, or push.\n\nFlags from the user: {flags}     # e.g. \"--clear\", \"--list\", or \"\" (none)\nFree-text intent (optional): {intent}\n\n## Step 1 — Snapshot state (parallel batch, single message, multiple Bash calls)\n\n- `git rev-parse --abbrev-ref HEAD`               # current branch\n- `git status --porcelain=v1`                     # never -uall\n- `git log -10 --oneline`                         # commit window\n- `git rev-list --count main..HEAD 2>\u002Fdev\u002Fnull`   # commits ahead of main\n- `git diff --stat main...HEAD 2>\u002Fdev\u002Fnull`       # changed-files summary\n- `gh pr view --json number,state,baseRefName,headRefName 2>\u002Fdev\u002Fnull` # open PR for branch\n- `cat .claude\u002Fstate\u002Fcurrent_plan.md 2>\u002Fdev\u002Fnull`                           # pointer to active plan (primary signal)\n- `ls -t .agents\u002Fplans\u002F*.md 2>\u002Fdev\u002Fnull | head -10`              # recent plans (mtime)\n- `ls -t .agents\u002Fplans\u002Fmeta-plan_*.md 2>\u002Fdev\u002Fnull | head -5`     # recent meta-plans\n\n## Step 2 — Resolve active plan (Status block is primary signal)\n\nResolution order — first match wins. **Steps 1+2 are read-only fast paths. Step 3 is the state-fixer fallback that prompts the user and persists their answer to `.claude\u002Fstate\u002Fcurrent_plan.md` so future invocations land in step 1.**\n\n### Step 2a — `.claude\u002Fstate\u002Fcurrent_plan.md` pointer (preferred)\n\nIf `current_plan.md` exists, parse it for the `**Plan:**` line. If the referenced plan file exists and has a `## Status` block (grep `^## Status$` to next `^## ` heading), use the block's `Active phase` + `Step` fields verbatim. Done — go to Step 3.\n\n### Step 2b — Status block on most-recent plan\n\nIf `current_plan.md` absent, glob `.agents\u002Fplans\u002Fplan_*.md` (mtime desc). For each, grep for `## Status` block. First match wins — use its fields. Don't read the whole plan, just the first 30 lines.\n\n### Step 2c — Inferred-state fallback (state-fixer)\n\nIf steps 2a+2b both fail (legacy plan with no Status block, or stale state, or no plan file matches), fall back to **commit-subject heuristic**:\n\n1. Identify newest plan via mtime: `ls -t .agents\u002Fplans\u002Fplan_*.md`\n2. Grep `^## Phase` headers in that plan to enumerate phases\n3. Grep commit subjects in `main..HEAD` (`git log --format=%s main..HEAD`)\n4. Cross-reference: a phase is **done** if its title (or commit-subject keywords from it, e.g. \"load_exclusive\", \"grimoire lock\") appears in any commit subject. The first phase with no commit-subject match is **active**.\n5. **Prompt the user via `AskUserQuestion`** with the inferred state:\n\n       Inferred from commits + plan headers: plan=\u003CX>, phase=\u003CN> \"\u003Ctitle>\" active.\n       Confirm and persist to .claude\u002Fstate\u002Fcurrent_plan.md?\n\n   Options:\n   - **Confirm + persist** (default) — write `current_plan.md`, inject `## Status` block at top of plan file (after H1, before next heading). Future `\u002Fnext` invocations land in Step 2a.\n   - **Different phase** — user names phase number; same persistence.\n   - **No active plan** — write nothing, treat as Step 2d.\n\nIf `--list` flag set, do not prompt — present the inferred phase as candidate #1 and let user pick.\n\n### Step 2d — No active plan (valid empty state)\n\nIf user picks \"No active plan\" in Step 2c, or no plan files exist at all:\n\n- `commits-ahead-of-main = 0` AND clean tree → say so explicitly, suggest browsing open issues\u002FPRs (`gh issue list --state open --limit 5` or `\u002Fhex-plan \"\u003Ctask>\"` for new work)\n- `commits-ahead > 0` AND no plan → suggest `\u002Fhex-plan` to capture the in-flight work as a tracked plan, OR `\u002Fcommit` if there's a Checkpoint\n- Right after `\u002Ffinalize` clears `current_plan.md` → \"feature just landed, pick next PR\u002Fissue\"\n\nEmpty state is a **valid outcome**, not an error.\n\n## Step 3 — Classify the phase\n\nApply the first matching row. **Status-block fields beat heuristic table** when both apply.\n\n| Observed state                                                          | Suggest                                       |\n|---|---|\n| Status `Step: \u002Fhex-execute → ...` OR plan-approved OR phase done with next phase | `\u002Fhex-execute \u003Ctier> \u003Cplan-path>` |\n| Status `Step: \u002Fhex-review → round N`                                  | `\u002Fhex-review` (continue same round)         |\n| Status all phases done, dirty tree                                      | `\u002Fcommit`                                     |\n| Status all phases done, clean tree, no PR                               | `\u002Fhex-review` then `\u002Ffinalize`              |\n| On `main`                                                               | `\u002Fhex-plan \"\u003Ctask>\"` — never work on main   |\n| Branch exists, no plan, no commits ahead                                | `\u002Fhex-plan \"\u003Ctask>\"` (ask if no task named) |\n| Meta-plan exists, no `plan_*.md` yet                                    | Awaiting plan approval — print `\u002Fhex-plan`  |\n| Plan exists, 0 commits ahead, clean tree                                | `\u002Fhex-execute \u003Cplan-path>`                  |\n| Plan exists, commits ahead, dirty tree                                  | `\u002Fcommit`                                     |\n| PR exists (state OPEN)                                                  | `\u002Fhex-review #\u003CN>`                          |\n| Worktree branch (goat\u002Fevelynn\u002Fsion\u002Fsoraka), multiple Checkpoints, clean | `\u002Ffinalize` (only if Status confirms phases done) |\n| HEAD subject == `Checkpoint`, clean tree                                | `\u002Fcommit` (will draft + amend)                |\n\nWhen two rows tie, pick the earlier (correctness > convenience). **Never suggest `\u002Ffinalize` without confirming via Status block that no phases are still active** — that was the regression this skill is fixing.\n\n## Step 4 — Synthesize the command\n\nRules — strictly enforced when `--clear` is set; recommended otherwise:\n\n- Use absolute paths to plan artifacts (`.agents\u002Fplans\u002Fplan_X.md`).\n- Inline tier and overlay flags from the plan's handoff block when known (`\u002Fhex-execute max .agents\u002Fplans\u002Fplan_X.md`).\n- For `\u002Fhex-review`, prefer a PR number (`#61`) when an open PR exists; otherwise the branch name.\n- For `\u002Fcommit` and `\u002Ffinalize`, no arguments — they snapshot state themselves.\n- For `\u002Fhex-plan`, include a 1-line task description in quotes.\n- NEVER reference \"the previous work\" \u002F \"what we just discussed\" \u002F \"the bug from earlier\".\n- NEVER suggest a slash command that is not present in `.claude\u002Frules.md` \"Skills by task topic\" — read that table if uncertain.\n\n## Step 5 — Reply\n\nOutput exactly one of these three formats. No preamble, no signature.\n\nDEFAULT (no flag):\n~~~\nNext:\n    \u003Cslash-command>\n\nWhy:    \u003Cone-line reason>\nLooked at: \u003Cone-line list of state checks>\n~~~\n\n--clear:\n~~~\nNext (handover — context will be cleared):\n    \u002Fclear\n    \u003Cself-contained slash-command>\n\nWhy:    \u003Cone-line reason>\nLooked at: \u003Cone-line list of state checks>\nNote: command is self-contained — safe to run after \u002Fclear.\n~~~\n\n--list:\n~~~\nCandidates:\n  1. \u003Cslash-command>\n       \u003Cone-line rationale, why this is #1>\n  2. \u003Cslash-command>\n       \u003Cone-line rationale>\n  3. \u003Cslash-command>\n       \u003Cone-line rationale>\n~~~\n\n## Edge cases\n\n- No git repo → reply: \"not in a git repo — nothing to suggest\". Exit.\n- Detached HEAD → suggest `git checkout \u003Cbranch>` first; do not propose a hex command.\n- `gh` unavailable \u002F unauthenticated → skip PR detection, fall through to branch baseline.\n- Multiple plans with mtime within 1h → list them, ask the user to pick (still one Candidates block).\n- Skill list expanded recently → re-read `.claude\u002Frules.md` \"Skills by task topic\" before mapping. Never invent a skill. The `\u002Fhex-*` commands listed there are user-level and will not appear under `.claude\u002Fskills\u002F` — that is expected.\n\nKeep your reply under 12 lines. The user will copy-paste from your output directly.\n```\n\n### 3. Print subagent's output verbatim\n\nSubagent already formatted output. Do not rewrap, summarise, comment. Print exactly what returned. If `--clear` set, subagent output already include `\u002Fclear` line — do not add another.\n\nIf subagent report hard error (no git repo, detached HEAD, etc.), print message and stop.\n\n## Constraints\n\n- NEVER edit code, commit, push.\n- May write `.claude\u002Fstate\u002Fcurrent_plan.md` and inject `## Status` block into a plan **only via user-confirmed Step 2c state-fixer path**. Never silently mutate state. The user's `AskUserQuestion` answer is consent.\n- NEVER run inspection commands in main session — subagent's job. Skip delegation = defeat context-isolation.\n- NEVER suggest a slash command that is neither in `.claude\u002Fskills\u002F` nor a hex command (`\u002Fhex-init`, `\u002Fhex-plan`, `\u002Fhex-execute`, `\u002Fhex-review`, `\u002Fhex-architect`) — hex ships at user level, so its absence from `.claude\u002Fskills\u002F` is expected, not an invented skill. Subagent cross-checks `.claude\u002Frules.md`; main session no second-guess.\n- ALWAYS spawn exactly one subagent. No parallelism — only one question asked.\n- NEVER suggest `\u002Ffinalize` without confirming via Status block that all phases are marked done. (See Step 3 table — premature finalize was the regression that motivated this skill's redesign.)\n\n## References\n\n- `.claude\u002Frules.md` — \"Skills by task topic\" (authoritative skill list — subagent reads it)\n- `.claude\u002Frules\u002Fworkflow-intent.md` — work-type router\n- `.claude\u002Frules\u002Fworkflow-git.md` — branching, two-phase commit\u002Ffinalize\n- `.claude\u002Frules\u002Fmeta-ai-config.md` \"Plan Status Protocol\" — schema + per-skill mutation table; `.claude\u002Fstate\u002F` tree is gitignored (per-worktree)\n- `\u002Fhex-execute`'s closing handoff — the pattern this skill emulates","# \u002Fnext — Suggest the Next Slash Command\n\nLook at current state (branch, working tree, plans, recent commits, open PR). Print exactly **one** slash command user copy-paste to continue. State inspection delegated to **sonnet subagent** so output stay compact, main session context not flooded with `git log` \u002F `gh pr view` \u002F plan-file output.\n\n**Primarily read-only.** Becomes a one-shot **state-fixer** when no `## Status` block found and no `.claude\u002Fstate\u002Fcurrent_plan.md` pointer — prompts user to confirm inferred state, then writes `current_plan.md` + injects Status block into plan so future invocations land in the fast path. Never edits code, never commits, never pushes.\n\n## Flags\n\n- `--clear` — tell subagent make suggested command **fully self-contained** (absolute paths, no \"the work we just did\" refs), prepend `\u002Fclear` so future session with no memory run cleanly.\n- `--list` — subagent return top 3 candidates with one-line rationale each.\n- (no flag) — one self-contained command + one-line \"why\" + one-line \"what was checked\".\n\n## Workflow\n\n### 1. Parse flags\n\nStrip `--clear` and `--list` from argument string. Leftover = free-text intent (rare; e.g. `\u002Fnext --clear after fixing the bug`).\n\n### 2. Delegate to sonnet subagent\n\nSpawn **one** subagent (`Agent` tool, `subagent_type: general-purpose`, `model: sonnet`) with prompt below. Do NOT run inspection commands yourself — point of delegation = keep main context clean.\n\n**Subagent prompt template** (fill `{flags}` and `{intent}` placeholders):\n\n```\nYou are inspecting Grimoire project state to suggest exactly ONE next slash command for the user to type. You are read-only — never edit, commit, or push.\n\nFlags from the user: {flags}     # e.g. \"--clear\", \"--list\", or \"\" (none)\nFree-text intent (optional): {intent}\n\n## Step 1 — Snapshot state (parallel batch, single message, multiple Bash calls)\n\n- `git rev-parse --abbrev-ref HEAD`               # current branch\n- `git status --porcelain=v1`                     # never -uall\n- `git log -10 --oneline`                         # commit window\n- `git rev-list --count main..HEAD 2>\u002Fdev\u002Fnull`   # commits ahead of main\n- `git diff --stat main...HEAD 2>\u002Fdev\u002Fnull`       # changed-files summary\n- `gh pr view --json number,state,baseRefName,headRefName 2>\u002Fdev\u002Fnull` # open PR for branch\n- `cat .claude\u002Fstate\u002Fcurrent_plan.md 2>\u002Fdev\u002Fnull`                           # pointer to active plan (primary signal)\n- `ls -t .agents\u002Fplans\u002F*.md 2>\u002Fdev\u002Fnull | head -10`              # recent plans (mtime)\n- `ls -t .agents\u002Fplans\u002Fmeta-plan_*.md 2>\u002Fdev\u002Fnull | head -5`     # recent meta-plans\n\n## Step 2 — Resolve active plan (Status block is primary signal)\n\nResolution order — first match wins. **Steps 1+2 are read-only fast paths. Step 3 is the state-fixer fallback that prompts the user and persists their answer to `.claude\u002Fstate\u002Fcurrent_plan.md` so future invocations land in step 1.**\n\n### Step 2a — `.claude\u002Fstate\u002Fcurrent_plan.md` pointer (preferred)\n\nIf `current_plan.md` exists, parse it for the `**Plan:**` line. If the referenced plan file exists and has a `## Status` block (grep `^## Status$` to next `^## ` heading), use the block's `Active phase` + `Step` fields verbatim. Done — go to Step 3.\n\n### Step 2b — Status block on most-recent plan\n\nIf `current_plan.md` absent, glob `.agents\u002Fplans\u002Fplan_*.md` (mtime desc). For each, grep for `## Status` block. First match wins — use its fields. Don't read the whole plan, just the first 30 lines.\n\n### Step 2c — Inferred-state fallback (state-fixer)\n\nIf steps 2a+2b both fail (legacy plan with no Status block, or stale state, or no plan file matches), fall back to **commit-subject heuristic**:\n\n1. Identify newest plan via mtime: `ls -t .agents\u002Fplans\u002Fplan_*.md`\n2. Grep `^## Phase` headers in that plan to enumerate phases\n3. Grep commit subjects in `main..HEAD` (`git log --format=%s main..HEAD`)\n4. Cross-reference: a phase is **done** if its title (or commit-subject keywords from it, e.g. \"load_exclusive\", \"grimoire lock\") appears in any commit subject. The first phase with no commit-subject match is **active**.\n5. **Prompt the user via `AskUserQuestion`** with the inferred state:\n\n       Inferred from commits + plan headers: plan=\u003CX>, phase=\u003CN> \"\u003Ctitle>\" active.\n       Confirm and persist to .claude\u002Fstate\u002Fcurrent_plan.md?\n\n   Options:\n   - **Confirm + persist** (default) — write `current_plan.md`, inject `## Status` block at top of plan file (after H1, before next heading). Future `\u002Fnext` invocations land in Step 2a.\n   - **Different phase** — user names phase number; same persistence.\n   - **No active plan** — write nothing, treat as Step 2d.\n\nIf `--list` flag set, do not prompt — present the inferred phase as candidate #1 and let user pick.\n\n### Step 2d — No active plan (valid empty state)\n\nIf user picks \"No active plan\" in Step 2c, or no plan files exist at all:\n\n- `commits-ahead-of-main = 0` AND clean tree → say so explicitly, suggest browsing open issues\u002FPRs (`gh issue list --state open --limit 5` or `\u002Fhex-plan \"\u003Ctask>\"` for new work)\n- `commits-ahead > 0` AND no plan → suggest `\u002Fhex-plan` to capture the in-flight work as a tracked plan, OR `\u002Fcommit` if there's a Checkpoint\n- Right after `\u002Ffinalize` clears `current_plan.md` → \"feature just landed, pick next PR\u002Fissue\"\n\nEmpty state is a **valid outcome**, not an error.\n\n## Step 3 — Classify the phase\n\nApply the first matching row. **Status-block fields beat heuristic table** when both apply.\n\n| Observed state                                                          | Suggest                                       |\n|---|---|\n| Status `Step: \u002Fhex-execute → ...` OR plan-approved OR phase done with next phase | `\u002Fhex-execute \u003Ctier> \u003Cplan-path>` |\n| Status `Step: \u002Fhex-review → round N`                                  | `\u002Fhex-review` (continue same round)         |\n| Status all phases done, dirty tree                                      | `\u002Fcommit`                                     |\n| Status all phases done, clean tree, no PR                               | `\u002Fhex-review` then `\u002Ffinalize`              |\n| On `main`                                                               | `\u002Fhex-plan \"\u003Ctask>\"` — never work on main   |\n| Branch exists, no plan, no commits ahead                                | `\u002Fhex-plan \"\u003Ctask>\"` (ask if no task named) |\n| Meta-plan exists, no `plan_*.md` yet                                    | Awaiting plan approval — print `\u002Fhex-plan`  |\n| Plan exists, 0 commits ahead, clean tree                                | `\u002Fhex-execute \u003Cplan-path>`                  |\n| Plan exists, commits ahead, dirty tree                                  | `\u002Fcommit`                                     |\n| PR exists (state OPEN)                                                  | `\u002Fhex-review #\u003CN>`                          |\n| Worktree branch (goat\u002Fevelynn\u002Fsion\u002Fsoraka), multiple Checkpoints, clean | `\u002Ffinalize` (only if Status confirms phases done) |\n| HEAD subject == `Checkpoint`, clean tree                                | `\u002Fcommit` (will draft + amend)                |\n\nWhen two rows tie, pick the earlier (correctness > convenience). **Never suggest `\u002Ffinalize` without confirming via Status block that no phases are still active** — that was the regression this skill is fixing.\n\n## Step 4 — Synthesize the command\n\nRules — strictly enforced when `--clear` is set; recommended otherwise:\n\n- Use absolute paths to plan artifacts (`.agents\u002Fplans\u002Fplan_X.md`).\n- Inline tier and overlay flags from the plan's handoff block when known (`\u002Fhex-execute max .agents\u002Fplans\u002Fplan_X.md`).\n- For `\u002Fhex-review`, prefer a PR number (`#61`) when an open PR exists; otherwise the branch name.\n- For `\u002Fcommit` and `\u002Ffinalize`, no arguments — they snapshot state themselves.\n- For `\u002Fhex-plan`, include a 1-line task description in quotes.\n- NEVER reference \"the previous work\" \u002F \"what we just discussed\" \u002F \"the bug from earlier\".\n- NEVER suggest a slash command that is not present in `.claude\u002Frules.md` \"Skills by task topic\" — read that table if uncertain.\n\n## Step 5 — Reply\n\nOutput exactly one of these three formats. No preamble, no signature.\n\nDEFAULT (no flag):\n~~~\nNext:\n    \u003Cslash-command>\n\nWhy:    \u003Cone-line reason>\nLooked at: \u003Cone-line list of state checks>\n~~~\n\n--clear:\n~~~\nNext (handover — context will be cleared):\n    \u002Fclear\n    \u003Cself-contained slash-command>\n\nWhy:    \u003Cone-line reason>\nLooked at: \u003Cone-line list of state checks>\nNote: command is self-contained — safe to run after \u002Fclear.\n~~~\n\n--list:\n~~~\nCandidates:\n  1. \u003Cslash-command>\n       \u003Cone-line rationale, why this is #1>\n  2. \u003Cslash-command>\n       \u003Cone-line rationale>\n  3. \u003Cslash-command>\n       \u003Cone-line rationale>\n~~~\n\n## Edge cases\n\n- No git repo → reply: \"not in a git repo — nothing to suggest\". Exit.\n- Detached HEAD → suggest `git checkout \u003Cbranch>` first; do not propose a hex command.\n- `gh` unavailable \u002F unauthenticated → skip PR detection, fall through to branch baseline.\n- Multiple plans with mtime within 1h → list them, ask the user to pick (still one Candidates block).\n- Skill list expanded recently → re-read `.claude\u002Frules.md` \"Skills by task topic\" before mapping. Never invent a skill. The `\u002Fhex-*` commands listed there are user-level and will not appear under `.claude\u002Fskills\u002F` — that is expected.\n\nKeep your reply under 12 lines. The user will copy-paste from your output directly.\n```\n\n### 3. Print subagent's output verbatim\n\nSubagent already formatted output. Do not rewrap, summarise, comment. Print exactly what returned. If `--clear` set, subagent output already include `\u002Fclear` line — do not add another.\n\nIf subagent report hard error (no git repo, detached HEAD, etc.), print message and stop.\n\n## Constraints\n\n- NEVER edit code, commit, push.\n- May write `.claude\u002Fstate\u002Fcurrent_plan.md` and inject `## Status` block into a plan **only via user-confirmed Step 2c state-fixer path**. Never silently mutate state. The user's `AskUserQuestion` answer is consent.\n- NEVER run inspection commands in main session — subagent's job. Skip delegation = defeat context-isolation.\n- NEVER suggest a slash command that is neither in `.claude\u002Fskills\u002F` nor a hex command (`\u002Fhex-init`, `\u002Fhex-plan`, `\u002Fhex-execute`, `\u002Fhex-review`, `\u002Fhex-architect`) — hex ships at user level, so its absence from `.claude\u002Fskills\u002F` is expected, not an invented skill. Subagent cross-checks `.claude\u002Frules.md`; main session no second-guess.\n- ALWAYS spawn exactly one subagent. No parallelism — only one question asked.\n- NEVER suggest `\u002Ffinalize` without confirming via Status block that all phases are marked done. (See Step 3 table — premature finalize was the regression that motivated this skill's redesign.)\n\n## References\n\n- `.claude\u002Frules.md` — \"Skills by task topic\" (authoritative skill list — subagent reads it)\n- `.claude\u002Frules\u002Fworkflow-intent.md` — work-type router\n- `.claude\u002Frules\u002Fworkflow-git.md` — branching, two-phase commit\u002Ffinalize\n- `.claude\u002Frules\u002Fmeta-ai-config.md` \"Plan Status Protocol\" — schema + per-skill mutation table; `.claude\u002Fstate\u002F` tree is gitignored (per-worktree)\n- `\u002Fhex-execute`'s closing handoff — the pattern this skill emulates",{"name":5,"triggers":23,"description":7,"argument-hint":29,"user-invocable":30,"disable-model-invocation":30},[24,25,26,27,28],"what's next","what is next","next step","next command","what now","[--clear | --list]",true,"main","ec018c25128e4b5b2e08679288e9e9153085830a",".claude\u002Fskills\u002Fnext",1,"https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire","https:\u002F\u002Fgrimoire.rs","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs","2026-11-14 09:47:24",{"source":8,"skillName":5,"commands":40,"prompt":50,"zipAvailable":30},[41,44,47],{"tool":42,"command":43},"npx","npx skills add grimoire-rs\u002Fgrimoire --skill next",{"tool":45,"command":46},"bunx","bunx skills add grimoire-rs\u002Fgrimoire --skill next",{"tool":48,"command":49},"pnpm","pnpm dlx skills add grimoire-rs\u002Fgrimoire --skill next","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fnext\u002FSKILL.md",[],[53,61,68,75,82,89],{"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":60},"01M78R7R0RQWY8BC0AQSB3E6NX","rca","Use when a bug ticket, GitHub issue or error report needs a root cause analysis before code is touched — \"is this a bug\", \"why does this fail\". Read-only. Gives a verdict (bug or not-a-bug kind), file:line evidence, ruled-out hypotheses, other affected sites, fix direction and size. A ticket, not a diff (hex-review).","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Frca\u002FSKILL.md",[],[],"2026-11-15 14:19:17",{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":60},"01M78R7JY3CY2XDNTB330GXTJK","fix","Use when a bug ticket or GitHub issue should go from report to an approved fix plan — \"plan a fix\", \"fix this issue\". Root-causes it via rca, then plans in plan mode with the failing regression test first. Routes to hex-plan instead when the fix is large or moves an interface contract.","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Ffix\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":18},"01M72XYVY9FGB38MZDJ0QFFMCC","p9-skill","Principle 9 differential fixture skill (v2).","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002Ftest\u002Ftools\u002Fp9_fixtures\u002Fv2\u002Fskills\u002Fp9-skill\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":18},"01M72XYT3PPTXEFWCR8H4GWKEG","p9-member","Bundle member skill for the Principle 9 fixture (v2).","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002Ftest\u002Ftools\u002Fp9_fixtures\u002Fv2\u002Fskills\u002Fp9-member\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":18},"01M72XYRWQWZ4BD0367QWG96TB","p9-hookish","Hook-free skill carrying an ignored hook.toml (v2).","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002Ftest\u002Ftools\u002Fp9_fixtures\u002Fv2\u002Fskills\u002Fp9-hookish\u002FSKILL.md",[],[],{"skillNo":90,"name":70,"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":94,"publishedAt":18},"01M72XYPDXXK5C6ZDD5BS2N6HS","Principle 9 differential fixture skill (v1).","https:\u002F\u002Fgithub.com\u002Fgrimoire-rs\u002Fgrimoire\u002Fblob\u002Fmain\u002Ftest\u002Ftools\u002Fp9_fixtures\u002Fv1\u002Fskills\u002Fp9-skill\u002FSKILL.md",[],[]]