session-close
Skill0 starsChanged 4 months agoMITView on GitHub
Use when ending a work session — wraps up with director reflections, consistency gate, /update ritual, session reflection, and session log. Fires on 'wrap up', 'close session', 'we're done', or on mechanical triggers (spec archived, ADR accepted, 5+ commits, phase gate pass).
- Commits and pushes
22 other projects use a version of this.
Adopted by
Most recently taken up by
- kimkon1011/orgiast-claude-rulestheir own version, first seen Nov 8, 2026
- paureis/pau-skillstheir own version, first seen Nov 8, 2026
- pankaj-mahaur/Agent-Daemontheir own version, first seen Nov 6, 2026
- NathanForster/claude-commandstheir own version, first seen Nov 1, 2026
- ngmeyer/skillstheir own version, first seen Nov 1, 2026
- albertomarturelo/siitheir own version, first seen Nov 1, 2026
Used with
Projects with it also use session-open and handoff.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
vishnujchandranSkillsession-open
Restart-safe session starter that restores context, validates readiness, and creates a focused execution plan for the current development session.
Also in 4 projects that use this
MicrosoftSkillhandoff
Compact the current conversation into a handoff document for another agent to pick up.
Also in 3 projects that use this
Other versions22 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 22
- albertomarturelo/sii★ 18same name
- tornikebolokadze1-cyber/agent-foundation-kit★ 6same name
- ngmeyer/skills★ 4same name
- paureis/pau-skills★ 2same name
- PUNCKe/manuscript-context-integrity★ 2same name
- toper2525ua-eng/claude-longrun★ 2same name
- vardan6/trackline★ 2same name
- AN0099/toothpaste-kit★ 1same name
- CorvusWing/CorvusWing★ 1same name
- farky-tech/containers★ 1same name
12 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository68 other files it gives its agents.
- adoptSkill
- architecture-decisionSkill
- architecture-reviewSkill
- art-bibleSkill
- asset-specSkill
- brainstormingSkill
- bug-reportSkill
- bug-triageSkill
- code-reviewSkill
- consistency-checkSkill
- content-auditSkill
- create-architectureSkill
- create-control-manifestSkill
- create-epicsSkill
- create-storiesSkill
- day-one-patchSkill
- design-reviewSkill
- design-systemSkill
- dev-storySkill
- dispatching-parallel-agentsSkill
- estimateSkill
- executing-plansSkill
- finishing-a-development-branchSkill
- gate-checkSkill
- goal-add-minorSkill
- goal-checkSkill
- goal-setSkill
- godot-engineSkill
- helpSkill
- map-systemsSkill
- narrative-designerSkill
- project-stage-detectSkill
- propagate-design-changeSkill
- qa-planSkill
- quick-designSkill
- red-flag-scanSkill
- regression-suiteSkill
- retrospectiveSkill
- review-all-gddsSkill
- safety-pushSkill
- scope-checkSkill
- security-auditSkill
- setup-engineSkill
- sprint-planSkill
- startSkill
- story-doneSkill
- story-readinessSkill
- subagent-driven-developmentSkill
- systematic-debuggingSkill
- team-artSkill
- team-audioSkill
- team-combatSkill
- team-economySkill
- team-infraSkill
- team-levelSkill
- team-marketingSkill
- team-narrativeSkill
- team-polishSkill
- team-releaseSkill
- team-securitySkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
10 rules. 0 are common, 7 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 18 edited versions we read keep these.
7
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 36
After spec archival (→ docs/z-old/)
- Distinctive: line 38
After 5+ commits without an update
- Distinctive: line 39
After phase gate passes
- Distinctive: line 40
After team-* orchestrator completes
- Distinctive: line 42
When mechanical trigger fires, prompt: *"That's a natural checkpoint — want me to run the close ritual, or keep going?"*
- Distinctive: line 60
Commit as PARTIAL with explicit carry-over destination
- Very specific: line 91
Run python tools/consistency_check.py (12 checks — registry, doc stack, ADRs, specs, links, hooks).
- Very specific: line 92
If any design doc changed this session → propose /review-all-gdds
- Distinctive: line 93
Check routine outcomes (architecture-review drift, code-review findings)
- Distinctive: line 114
After git push confirmed, capture a brief close-out reflection:
“Common” is measured: kept by at least 6 in 10 of the 18 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Session Close
🌱 ShiningPlague-authored. No upstream version exists. Unified close protocol — every session ends the same way (gate → directors → consistency → /update → git → session reflection).
Unified close protocol. Every session ends the same way.
If an artifact named here is absent: say so plainly in one line, skip that step, and continue. Never invent the file to satisfy a checklist, and never fail a close because an optional artifact was never created.
Triggers
Natural (the assistant initiates)
- Designer says "wrap up" / "we're done" / "good session" / "let's close"
- Last task in a plan completes
- Brainstorm reaches consensus and spec saved
- Designer shifts topic away from current work
Mechanical (safety net)
- After spec archival (→ docs/z-old/)
- After ADR acceptance
- After 5+ commits without an update
- After phase gate passes
- After team-* orchestrator completes
When mechanical trigger fires, prompt: "That's a natural checkpoint — want me to run the close ritual, or keep going?"
Procedure
Step 0: Pre-Close Self-Check — GATE
MUST fire BOTH /goal-check AND /red-flag-scan BEFORE Step 1. Together they form the gate.
0a — /goal-check — validates each primary + minor goal against actual evidence (artifacts shipped, commits made, expected outcomes met). Surfaces unresolved issues from active-goals.json.
0b — /red-flag-scan — broader systemic audit: uncommitted work, doc stack drift, complexity drift (skill count, CLAUDE.md lines, orthogonal state axes), broken hooks/skills, technical debt accumulation, framework inconsistency.
Verdict colors:
- 🟢 GREEN (all PASS) → continue to Step 1
- 🟡 YELLOW (some PARTIAL / DROPPED with carry-over OK) → continue to Step 1 with explicit carry-over commitments written to active.md + sprint-status.yaml
- 🔴 RED (critical fail / unresolved blocker) → CLOSE PAUSES. Surface critical fails to designer. Designer chooses:
- Fix now (in-chat resolution) → re-run /goal-check after fix
- Formally DROP with rationale (write to active-goals.json + active.md)
- Commit as PARTIAL with explicit carry-over destination
- Defer close entirely (return to work, close later)
Hard-won rule: this step exists because without it, closes march on while critical issues linger — directors return CONCERNS verdicts and the close continues to Step 2 without designer resolution. The gate forces the pause.
Output of Step 0 gets written to the session log (Step 7) as the gate verdict + designer disposition table.
Step 1: Director Bookend Re-dispatch
Fire all 3 Tier 1 directors in parallel via Agent tool with CLOSE-REFLECTION prompts from .claude/docs/director-gates.md:
creative-director→ gate CLOSE-CDtechnical-director→ gate CLOSE-TDproducer→ gate CLOSE-PR
Pass each: workstream name + activity, decisions made, scope shipped. Collect all 3 reflections before proceeding.
Step 2: Synthesis — Reflection Record
Compose unified reflection from all 3 director outputs:
- Director key concerns
- Alignment / tensions between departments
- Shipped list
- Carryover items
- Producer's food-for-thought question
- Top risk + top opportunity
Step 3: Close Gate Sequence
Before update ritual:
- Run
python tools/consistency_check.py(12 checks — registry, doc stack, ADRs, specs, links, hooks). Exit 0 = no FAILs; WARNs are advisory and never block. - If any design doc changed this session → propose
/review-all-gdds - Check routine outcomes (architecture-review drift, code-review findings)
- If the runner exits non-zero → update proceeds but git push blocked until the named FAILs are resolved
Step 4: Unified Update Ritual
Fire /update (project-local skill). It handles:
- git status + recent commits
- consistency_check.py (auto-bumps last_full_audit on PASS)
- devlog.md append
- dev_diary.json (done_major, done_minor, close_note, close_ref)
- active.md full state handoff
- Workstream state file update
- Registry updates for modified systems
- Flag contradictions
Step 5: Git Checkpoint
Draft commit message. Prompt designer for approval. Never push without explicit go.
Step 6: Session Reflection
After git push confirmed, capture a brief close-out reflection:
- Summarise the session — what went well, what was challenging, what to carry forward
- Distil a 2-sentence reflection note for the dev_diary entry
- Fold the director reflections (Step 2) and the reflection note into the session log (Step 7)
Step 7: Session Log Auto-File
Write production/session-logs/YYYY-MM-DD-[workstream]-close.md:
- Full director reflections
- Shipped list + commits
- Consistency check results
- Workstream state snapshot
- Reflection note
- Next session priority
Reference from dev_diary via close_ref field.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

