wrapup

A final command for the fifth stage of a five-stage software evaluation workflow. It checks code, synchronizes documentation, and commits changes.

In plain words
What is it for?
Use it to run formatting, linting, type checks, tests, documentation updates, and a final commit for the embedeval project.
Why use it?
It gathers the required checks into one finishing step and prevents commits when a quality check fails.

Command for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/ecro/embedeval/wrapup
Clone the repo
git clone --depth 1 https://github.com/Ecro/embedeval

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,082 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01082
Opus 5 $0.00000 $0.00541
Sonnet 5 $0.00000 $0.00216
Haiku 4.5 $0.00000 $0.00108

Measured yesterday against content hash 132712edb629, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

wrapup scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/wrapup.md · 124 lines

How it starts

The opening of the file, as written. The whole thing — 124 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/wrapup — Finalize and commit

Final validation, docs sync, and commit. Step 5 of the embedeval 5-stage workflow. Reports to screen only — no WRAPUP/LEARNING document.

Communication

  • Direct, evidence-based.
  • Never commit on a red gate. Never push without explicit user confirmation.
  • Report what actually happened — don't paper over a skipped step.

Usage

/wrapup [task-slug]

Slug is optional — used in commit message if provided; inferred from the PLAN referenced by the diff otherwise.

Context

  • Repo: /home/noel/embedeval
  • PLAN (if slug given): plans/PLAN-<slug>.md
  • Pre-commit hook at .githooks/pre-commit mirrors the ruff subset. Activate per clone: git config core.hooksPath .githooks. Never skip with --no-verify.
  • Quality gates — MANDATORY, all must pass:
    • uv run ruff format --check src/
    • uv run ruff check src/
    • uv run mypy src/
    • uv run pytest tests/ -v --tb=short
  • Doc auto-sync — MANDATORY when cases/, src/, or tests/ changed:
    • uv run python scripts/sync_docs.py

Steps

  1. Preflight

    • git status --short — show uncommitted changes.
    • Confirm branch. Warn if on main directly without the agent-branch convention: agent/<slug>/<YYYY-MM-DD>.
  2. Full quality gate

    • Run all four gates sequentially. Stop on first failure; report exact error and exit. Do not commit.
  3. Doc auto-sync

    • If the diff (staged + unstaged + last commit vs HEAD~1) touches cases/, src/, or tests/:
      • Run uv run python scripts/sync_docs.py.
      • If README.md or docs/METHODOLOGY.md changed, stage them.
    • If the script prints "already up to date", no action.
  4. Self-review diff

    • Print git diff --stat for staged + unstaged.
    • Sanity-check against CLAUDE.md "Quality Gates" #1 (API contracts, redundant calls, edge cases).
    • Ask user to confirm before committing.
  5. Refresh test tracker (if cases/ changed)

    if git diff --name-only HEAD~1 -- cases/ | grep -q .; then
        uv run embedeval refresh-tracker --cases cases/ --results results/
    fi
    

Read the full file on GitHub · 124 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 124 lines · 0 tokens per session scan A 132712edb629

Subscribe to this mod's changes

wrapup is a command published in the GitHub repository Ecro/embedeval (11 stars, last pushed 26d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 1,082 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.