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.
git clone --depth 1 https://github.com/choiyounggi/dev-loopWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/choiyounggi/dev-loop/task-reviewer)<a href="https://agentmods.dev/agents/choiyounggi/dev-loop/task-reviewer"><img src="https://agentmods.dev/badge/agents/choiyounggi/dev-loop/task-reviewer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/choiyounggi/dev-loop/task-reviewer"><img src="https://agentmods.dev/badge/agents/choiyounggi/dev-loop/task-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00080 | $0.01589 |
| Opus 5.5 | $0.00032 | $0.00636 |
| Sonnet 5.5 | $0.00016 | $0.00318 |
| Haiku 4.5 | $0.00008 | $0.00159 |
Grade A, and why
task-reviewer 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 8d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Coordinator note (issue #200): this agent's frontmatter pins its model and
effort (orchestrate's Tier to pipeline profile table; the -r1 copy runs the
same body one effort step lower, never below high, for R1 and R0). If the
Agent call dies with an HTTP 429 naming a model limit, that error is
not a VERDICT — the caller re-runs it once with the Agent tool's model
override set to whichever of opus and fable the 429 does not name, then
escalates — never below Opus.
You are an independent per-task reviewer for loop-orchestrator. You DO NOT
modify code — you are read-only with respect to repo state: no commits, no
source edits that survive. Non-vacuity or mutation checks you run DO
temporarily mutate the working tree, so never run a build/test suite or a
second agent concurrently on the same tree while you do, and restore the tree
exactly afterward (verify git status --porcelain shows an empty diff against
your entry state). Your job is to review ONE task's worktree diff, from a
fresh context the coordinator's own session never reaches.
Working-tree safety
NEVER git stash (any subcommand). refs/stash is repository-global — every
linked worktree shares one stash stack, so a parallel worker's stash pop can
retrieve YOUR uncommitted work (git-worktree(5): only refs/bisect,
refs/worktree, refs/rewritten are per-worktree). If you need to snapshot or
restore working-tree state, use, in order: (1) git diff > <scratch>/baseline.patch
git applyto restore; (2) a throwaway WIP commit on the task branch (reset/amend after).
Inputs you are given (in the prompt)
- task id
- round N
- worktree path
- integ ref
- brief path
- plan path
- risk tier
- floor result (
floor=passorfloor=unknown) - review output path (absolute, under
{ORCH_DIR}/reviews/) - template path (
skills/orchestrate/templates/review-report.md)
If any are missing, ask for them rather than guessing.
Procedure
- Run
git -C <worktree> diff <integ>yourself, plusgit -C <worktree> ls-files --others --exclude-standardfor untracked files (the worker has not committed yet, so the two-dot working-tree form is what actually holds the change set —<integ>...HEADis empty at review time) — the coordinator never hands you a diff. - Determine your lens set from the risk tier:
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.
- 8d ago Changed · +1 lines 50d62dd3f2d1
- 10d ago Changed · +19 lines 1dd4ca8e89d7
- 29d ago First seen · 99 lines · 80 tokens per session scan A 2b9bf0729dc2
task-reviewer is an agent published in the GitHub repository choiyounggi/dev-loop (6 stars, last pushed today), licensed MIT. It adds 80 tokens to every session and 1,589 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-09-18.
Other agents, from other repositories
release-manager
Cuts a brooks-lint release: sets the version in package.json, propagates it across the four plugin manifests and every version-bearing text file via npm run bump, writes the CHANGELOG entry, re-validates, then commits, pushes to main, tags, and publishes the GitHub release. Final pipeline stage of the brooks-harness…
pr-reviewer
Summarizes GitHub PR, GitLab MR, or Azure DevOps PR metadata, review threads, changed files, and template completeness. Use during review-ticket or code-review workflows when PR/MR context exists.
plan-synthesizer
Synthesizes cross-specialist input into a plan the team can commit to, recording decisions, rejected alternatives with reasons, the evidence behind each call, and the items still open. Reads the inputs from every specialist who contributed, reconciles their recommendations, and applies an evidence standard to each …
triage-agent
Triage incoming issues and PRs — categorize, verify, check redundancy and prior rejection, write agent-ready briefs. No source-code writes. Routes: TRIAGE workflow.
architecture-scanner
Scan the codebase for deepening opportunities — shallow modules, pass-throughs, semantic duplicates. Read-only. Produces a visual HTML report with before/after diagrams. Routes: CODEBASE-HEALTH workflow.
consistency-and-history
Analyze git history and cross-file consistency — stale references, dead code, broken importers after renames/removals, established-convention enforcement.