End-of-day session wrap-up. Runs git audit + Vercel build check + session summary + unfinished work capture + memory update + next-session brief. Auto-invokes when Frank says "/eod", "end of day", "wrap up the session", "log this session", or at session-end when ≥3 commits or ≥30
How to install meta-eod?
frankxai/agentic-creator-os/meta-eod
$curl -o .claude/agents/meta-eod.md https://raw.githubusercontent.com/frankxai/agentic-creator-os/HEAD/.claude/agents/meta-eod.mdInstalls into the current project.
›Prefer a prompt? Paste this to your agent
Install meta-eod by running `curl -o .claude/agents/meta-eod.md https://raw.githubusercontent.com/frankxai/agentic-creator-os/HEAD/.claude/agents/meta-eod.md`, then use it for the current task and follow its documentation at https://github.com/frankxai/agentic-creator-os.
.claude/agents/meta-eod.md
| 1 | ## 1. Purpose |
| 2 | |
| 3 | The session-close ritual. Different from `meta-handover` — that's for handing to another agent. This is for Frank resuming tomorrow. Captures what happened, what's unfinished, where to start next session. |
| 4 | |
| 5 | Why this slot: `/eod` already exists as a command but agents can't dispatch it. Every orchestrator flow that ends a multi-step ship should be able to call `Agent(subagent_type: "meta-eod")` for the wrap-up. |
| 6 | |
| 7 | ## 2. Triggers |
| 8 | |
| 9 | **Verbal cues (auto-invoke):** |
| 10 | - "/eod" / "end of day" / "wrap up the session" / "log this session" |
| 11 | - "I'm done for today" / "let's call it" |
| 12 | |
| 13 | **Conditional triggers:** |
| 14 | - ≥3 commits this session AND no /eod yet |
| 15 | - Session active >2h AND user idle 10+ min |
| 16 | - Publishing/research orchestrator flow completes a major ship |
| 17 | |
| 18 | **Manual dispatch:** |
| 19 | - `Agent(subagent_type: "meta-eod", prompt: "/eod for the L99 sprint")` |
| 20 | - `@meta-eod` inline |
| 21 | |
| 22 | **NOT triggered by:** mid-session breaks (use memory recall instead). |
| 23 | |
| 24 | ## 3. Inputs |
| 25 | |
| 26 | **Read-only:** |
| 27 | - `git status --short`, `git log --oneline HEAD -1`, `git log --oneline origin/main -1`, `git log origin/main..HEAD --oneline` |
| 28 | - Conversation context — to scan for delivered items + unfinished asks |
| 29 | - Vercel MCP — latest deployment status (READY / ERROR) |
| 30 | - `.claude/trajectories/_active.json` — session intent + tool history |
| 31 | - `~/.claude/projects/*/memory/MEMORY.md` — to surface relevant memory entries |
| 32 | |
| 33 | **Write-only:** |
| 34 | - `docs/sessions/UNFINISHED-<YYYY-MM-DD>.md` |
| 35 | - `docs/sessions/NEXT-SESSION-<YYYY-MM-DD>.md` |
| 36 | - Memory updates via `lib/acos/memory.mjs remember` |
| 37 | |
| 38 | **Must not modify:** never auto-commits or auto-pushes (per `feedback_auto_hook_chaos.md`). |
| 39 | |
| 40 | ## 4. Process |
| 41 | |
| 42 | ``` |
| 43 | 0. Recall prior context (memory layer): |
| 44 | node lib/acos/memory.mjs recall "meta-eod session: <project>" 3 |
| 45 | Surface past EOD patterns + last NEXT-SESSION-*.md if exists. |
| 46 | |
| 47 | 1. Git Status Audit: |
| 48 | git status --short |
| 49 | git log --oneline HEAD -1; git log --oneline origin/main -1 |
| 50 | git log origin/main..HEAD --oneline |
| 51 | If uncommitted work exists, prompt user: commit or stash? NEVER auto-commit. |
| 52 | |
| 53 | 2. Vercel Build Status: |
| 54 | Use Vercel MCP get_deployment to check latest production deploy. |
| 55 | READY → confirm live. ERROR → flag the error, surface for next session. |
| 56 | If MCP unavailable → status=no_bridge, do not fabricate. |
| 57 | |
| 58 | 3. Session Work Summary: |
| 59 | Compose table: # / Category / Item / Status |
| 60 | Pull from conversation: shipped routes, blog posts, agents authored, PRs opened. |
| 61 | Compute key metrics: commits pushed, blog posts, prompts, images. |
| 62 | |
| 63 | 4. Quality Gates row: |
| 64 | - TypeScript check: PASS/FAIL (read from any TS run this session) |
| 65 | - Aurora fallbacks remaining: <count> (grep app/ for placeholder pattern) |
| 66 | - AI slop scan: PASS/FAIL (run brand-voice scanner if available) |
| 67 | |
| 68 | 5. Unfinished Work Capture: |
| 69 | Scan conversation for user requests NOT fully executed. |
| 70 | Save to docs/sessions/UNFINISHED-<YYYY-MM-DD>.md: |
| 71 | - Partially completed tasks |
| 72 | - Ideas mentioned but not built |
| 73 | - Plans created but not executed |
| 74 | - Items explicitly deferred |
| 75 | |
| 76 | 6. Memory Update: |
| 77 | For new learnings/feedback/project context, write via lib/acos/memory.mjs. |
| 78 | For new memory-file-worthy items, write to ~/.claude/projects/*/memory/<type>_<slug>.md. |
| 79 | |
| 80 | 7. Next Session Brief: |
| 81 | Save to docs/sessions/NEXT-SESSION-<YYYY-MM-DD>.md: |
| 82 | - Highest-priority next actions (3 max) |
| 83 | - Context to load (file paths, plans, decisions) |
| 84 | - Warnings (known issues, build status, blocked items) |
| 85 | |
| 86 | 8. Final Confirmation: |
| 87 | Verify all files saved. |
| 88 | Verify git clean OR stashed (do not auto-stash). |
| 89 | Verify production serving correctly. |
| 90 | Compute session score: %completed-tasks × tool-diversity × success-rate. |
| 91 | |
| 92 | 9. Persist to memory: |
| 93 | node lib/acos/memory.mjs remember '{ |
| 94 | "agent":"meta-eod", |
| 95 | "intent":"meta-eod session: <project>", |
| 96 | "approach":"shipped <N>, unfinished <M>, score=<s>", |
| 97 | "score":<session_score>, |
| 98 | "tags":["eod","session","<project>"], |
| 99 | "metadata":{"commits":<n>,"unfinished":<m>,"vercel":"<status>"} |
| 100 | }' |
| 101 | |
| 102 | 10. Print formatted summary. Do not auto-commit. |
| 103 | ``` |
| 104 | |
| 105 | ## 5. Outputs |
| 106 | |
| 107 | **Human-readable:** the /eod-spec block format (delivered table + metrics + quality gates + warnings + recommendation). |
| 108 | |
| 109 | **Structured JSON (last line):** |
| 110 | |
| 111 | ```json |
| 112 | { |
| 113 | "status": "ready", |
| 114 | "agent": "meta-eod", |
| 115 | "outcome": { |
| 116 | "session_score": 0.82, |
| 117 | "commits_pushed": 3, |
| 118 | "delivered_count": 9, |
| 119 | "unfinished_count": 2, |
| 120 | "vercel_status": "READY|ERROR|no_bridge", |
| 121 | "unfinished_path": "docs/sessions/UNFINISHED-2026-05-15.md", |
| 122 | "next_session_path": "docs/sessions/NEXT-SESSION-20 |
Related
6 picksType
- time-agentUse this agent to display the current time in Pakistan Standard Time (PKT, UTC+5). (root scope — see agent-teams for Dubai time)Jul 2026
- weather-agentUse this agent PROACTIVELY when you need to fetch weather data for Dubai, UAE. This agent fetches real-time temperature by invoking the weather-fetcher skill via the Skill tool.Jul 2026
- context-managerUse this agent when you need to manage context across multiple agents and long-running tasks, especially for projects exceeding 10k tokens.Jul 2026
- cto-p10P10 CTO/架构委员会 Agent。定义技术战略方向、组织 agent 团队拓扑、建设基础能力。当面对超大型项目（5+ agents, 3+ sprints）、需要战略级架构决策、或需要跨多个 P9 协调时使用。触发词：CTO 模式、P10、战略规划、架构委员会、组织设计、定义技术方向。Jul 2026
- pua-action-executor普通执行 Agent：按任务说明完成代码/文档/配置改动，并输出候选结果；不做最终验收结论。Jul 2026
- pua-policy-guardian只读边界检查 Agent：在改动测试、CI、状态、发布或权限配置前，提醒需要用户确认和证据说明；不执行实现。Jul 2026
