agentleFS
gutohuidagutohuida/AgentWeave/.claude/skills/daily-review/SKILL.md

daily-review

Skill4 starsChanged 2 months agoApache-2.0View on GitHub

Read the daily loop's review page, talk the proposed specs through with the operator, and record their decisions so tonight's build window can act on them. Publishes the review page as an Artifact, answers questions about any proposal, applies requested edits to openspec/changes/, and writes spec-queue/APPROVALS.md. Use when the operator says "today's review", "daily review", "review the specs", "what did the loop propose", "approve today's specs", or opens a session after the day window has run. This is the human step of the cycle in openspec/explorations/2026-09-01-a-daily-research-spec-and-build-loop.md; the two scheduled windows either side of it are unattended and cannot ask anything.

18 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use weekly-review.

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.
  • QuixiAISkill

    weekly-review

    Review the past week across goals, tasks, calendar, journal, and commitments, then propose the next week

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 11 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep the same artifact URL across days if the operator wants one running page.

    line 48
  2. Distinctive:

    When the operator wants a change different:

    line 65
  3. Distinctive:

    Commit it. Then tell the operator, in one short list, exactly what tonight's window will build — including the backlog items that come …

    line 96
  4. Distinctive:

    Check git branch --merged master for the cycle branch.

    line 104
  5. Distinctive:

    Never mark a task complete on the strength of a plan existing.

    line 113
  6. Distinctive:

    Do not touch port 8000. 8010 is the trial Hub.

    line 114

“Common” is measured: kept by at least 6 in 10 of the 11 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

The daily loop's one human step. Everything either side of it runs unattended and cannot ask a question, so this is where judgement enters.

Contract for every file named here: spec-queue/README.md. Playbooks for the windows either side: .claude/loops/.

Step 1 — Find out what happened, before opening anything

powershell -NoProfile -Command "Get-Date -Format 'yyyy-MM-dd'"   # Git Bash date is skewed here
ls -t spec-queue/review/*.html | head -3
git branch --show-current
git log --oneline -15
git status --short

Then read, in this order:

  1. The newest review page in spec-queue/review/. If it is not today's, say so in your first sentence — the operator is about to make decisions and needs to know they are about a day they may not remember.
  2. The bottom of the newest .claude/autonomous/*-day-log.md, for what the window actually did versus what the page claims.
  3. spec-queue/APPROVALS.md — is there already a section for today? Someone may have started.

If there is no review page at all, do not improvise one. Report what you found — whether the day window ran, what its log says, whether AgentWeaveResearch produced a file in ~/.claude/routines/agentweave-research/out/ — and stop. A review page invented here has not been through the rounds and must not reach APPROVALS.md.

Step 2 — Publish it

Publish the review page as an Artifact and give the operator the link. This is the reason the human step exists in an interactive session at all: headless claude -p has no Artifact tool (verified 2026-08-28; no CLI flag provisions one), so the window that wrote the page could not publish it.

The page is already self-contained and theme-aware. Strip the outer <!DOCTYPE>/<html>/<head>/<body> wrapper before publishing — the Artifact runtime supplies its own skeleton and would nest them — but leave the file on disk intact, since it is also meant to open in a browser.

Keep the same artifact URL across days if the operator wants one running page; use a fresh one per day if they would rather have the history. Ask once, then stop asking.

Step 3 — Talk it through

This is the part that cannot be automated, so do not rush it into a checklist.

For each proposed change, the operator needs enough to decide and no more: what breaks today, what the change does about it, what it touches, and its tier, its acceptance drive, and what the review round changed. That last one is the tell. A change where two independent re-derivations found nothing is either genuinely clean or was not really re-derived, and the operator is entitled to know which the page claims.

Answer questions by reading the actual proposal and the actual code — openspec/changes/<name>/ and the files it names. Do not answer from the review page alone; it is a summary written by a process that was optimising for brevity.

When the operator wants a change different:

  • Small correction — apply it to openspec/changes/<name>/ now, re-run openspec validate --strict <name>, and mark the row APPROVED. Say in the note what you changed.
  • The argument is wrong — mark it REVISING with what needs re-deriving. Tomorrow's day window takes another round. Do not rewrite an argument here and then approve your own rewrite; that is the round discipline collapsed to one round, performed by the person who is meant to be checking it.
  • Not now — REJECTED, with the reason. The reason matters: without one it will be re-proposed.

Step 4 — Write the decisions down

Append today's section to spec-queue/APPROVALS.md, newest day first, using the format in spec-queue/README.md. The status token is the authority; there is no checkbox.

## <today>

- APPROVED  <change-name>   note
- REVISING  <change-name>   what needs re-deriving
- REJECTED  <change-name>   why

Offer the review page's Tier-0 fix list and write the operator's picks as one APPROVED-FIXES: F<n>, F<n> line. Add ORDER: only if the operator asked for a different order tonight. Add NOTHING TONIGHT if they want the window to stand down — it stops before spending a model invocation.

Leave undecided changes out. Absence is not rejection: a row-less change stays in openspec/changes/ and comes back in tomorrow's page. Inventing a decision the operator did not make is the one failure this step cannot recover from, because the night window treats this file as authoritative and nobody is awake to contradict it.

Commit it. Then tell the operator, in one short list, exactly what tonight's window will build — including the backlog items that come first by default, so the order is never a surprise in the morning.

Step 5 — The merge question

The cycle branch is merged by the operator, awake. Never automatically, and never by a window.

Check git branch --merged master for the cycle branch. If it is unmerged and spans more than a day, raise it — an accumulating diff that nobody merges is worse than no loop, because it eventually grows past reading. Offer the merge; do not perform it unasked.

Limits

  • Push, do not open PRs.
  • Stage explicit paths, never git add -A.
  • openspec validate --strict <name> must pass before any row is marked APPROVED.
  • Never mark a task complete on the strength of a plan existing.
  • Do not touch port 8000. 8010 is the trial Hub; the windows' drive Hub uses a port chosen that night, which the window's log records.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.