Write the session-metrics card for the branch named in $ARGUMENTS, or the current branch if it is empty. Then stop.

What this run must produce

One thing: the card. .claude/metrics/<branch-slug>.json, carrying this branch's pull request number, issue number and issue labels — committed, pushed, and rendered onto the pull-request body as a <details> block.

Nothing else. No RETRO.md, no findings, no changes to the skills, CLAUDE.md, the wiki, the tests or the brief.

When to run this instead of retro

retro is the full procedure: the card, then reading what the session cost against the complexity declared before it started, then turning that into concrete changes. That reading is the point of the exercise, and this skill throws it away.

Run this one when the session has no room left to do it properly — a long run close to its context or token budget, where the choice is between a card and nothing. Run retro in every other case.

A card written here is complete and permanent: it holds the same measurements a full run would have written, and retro can read it later from any session. The analysis is what is deferred, not the record.

Where this sits

new-feat, implement, prep-pr, then this. Under orchestrate, once per task, after that task's pull request is open and before the shepherd watches it.

After prep-pr and not inside it, because prep-pr dispatches three review agents: a card written before they finish measures building the change rather than shipping it.

Before Step 1 — what can actually run

Three answers, once, exactly as prep-pr establishes them:

git ls-remote --exit-code origin HEAD >/dev/null 2>&1 && echo "network: yes" || echo "network: no"
command -v gh >/dev/null && gh auth status >/dev/null 2>&1 && echo "gh: yes" || echo "gh: no"
ls ~/.claude/projects >/dev/null 2>&1 && echo "transcripts: yes" || echo "transcripts: no"

If prep-pr ran in this same session it already has these three answers. Read them out of its report rather than probing again.

transcripts: no governs what the collector can write. It reads ~/.claude/projects; with that directory absent, the card it writes now carries the diff and zero spend — true, but able to answer nothing about cost. Write it anyway, and say in the final message that the spend figure is not a measurement. Never present a zero-spend card as a cheap session.

A card already in .claude/metrics/ is a record, not collector output to be distrusted. Its spend was measured when the transcripts still existed and stays true after they are gone — that is the entire reason the file is committed rather than recomputed.

Step 1 — Write the card

BRANCH=$(git branch --show-current)
PR=$(gh pr view --json number --jq .number)

Then the two runs, which do different jobs:

# 1. The committed card. `--resolve-issue` finds the pull request, its first
#    linked issue and that issue's `complexity:` label for itself.
bun run --cwd tools/pr-metrics card -- --resolve-issue

# 2. The `<details>` block for the pull-request body, on stdout, warnings excluded.
bun run --cwd tools/pr-metrics card -- --resolve-issue --format markdown > /tmp/card.md

Four rules about that pair:

• Let --resolve-issue do the lookup; never hand-roll it in the shell. .closingIssuesReferences[0].number drops the repository, so a follow-up gh issue view <n> --repo xchromo/osn succeeds against a different repository's issue of the same number and writes a stranger's complexity: label in as this branch's denominator. resolveIdentity in tools/pr-metrics/index.ts compares the reference's own repository.
• Labels are read only from an issue in this repository. Most work here closes a finding in the private xchromo/osn-tracker, and that issue's severity:/area: labels must never reach a card committed to a public repository. The flag withholds them and records complexity.method: "not-fetched", which is not "none" — a rating may well exist there and nobody looked. "lookup-failed" is its third answer, for a call that did not land.
• Pass explicit flags only where you know better than gh does. --pr, --issue, --issue-labels and --complexity all win over the lookup, and passing one turns the rest of it off. The usual reason is a branch whose pull request is not open yet.
• Commit the JSON with the branch. ~/.claude/projects is local, unversioned and dies with a remote container. The committed file is the record; an uncommitted one is nothing. Stage the one card, never the directory: .claude/metrics/ holds every other branch's card too, and a git add of the whole of it sweeps in anything a concurrent session left there.

git add ".claude/metrics/$(echo "$BRANCH" | tr -c 'a-zA-Z0-9._-' '-').json"
git commit -m "chore: session-metrics card for $BRANCH"
git push

Step 2 — Render it onto the pull request

gh pr view "$PR" --json body --jq .body > /tmp/body.md
cat /tmp/card.md >> /tmp/body.md
gh pr edit "$PR" --body-file /tmp/body.md

Read the body out and append to it — never compose a new one. The five sections prep-pr wrote are the pull request's contract, and an edit that rebuilds the body from memory loses them.

Check the card is not already on the body before appending — a re-run would otherwise stack a second block. Match the summary line, not the tag: a body that describes what this step does mentions <details> in its own prose, and a bare grep -c '<details>' then reports a block that is not there.

gh pr view "$PR" --json body --jq .body | grep -c '^<details><summary>Session metrics'

The block is a <details>, never a ## section: prep-pr's body shape check permits exactly five top-level headings, and appending a sixth fails a body that is otherwise correct.

When a step here cannot run, record which and carry on — a missing card never blocks anything. No gh: write the card with whatever identity you know and leave the block in a file, naming that file in the final message. No network: commit the card locally; it pushes with the next push.

The SessionEnd hook is a fallback, not a second writer. It runs card -- --if-absent --resolve-issue, so it writes a card for a branch that has none and never touches one this skill committed.

A card with pr.number: null is one the hook wrote before a pull request existed. Its spend and interaction numbers are as good as any other card's — only its identity is missing, and Step 1 fills that in where it can run.

Finish

Report the card's path, the pull request it names, the issue and complexity.declared it resolved, and spend.usd_equivalent. Say plainly that the analysis half was not run and that retro reads this card later from any session.

Do not offer a summary of what the session cost, and do not volunteer findings. Someone who wants those runs retro.