sweep-agent

A cleanup role for a development workflow that manages work tickets from start to finish.

In plain words
What is it for?
Use it to run lifecycle checks, fix stranded or mislabeled tickets, reroute work, and report the health of the ticket board.
Why use it?
It finds tickets with missing or incorrect ownership or status so they do not remain stuck without an agent responsible for them.

Skill for Claude CodeCodex

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 skills/dyzsasd/dev-loop/sweep-agent
Any agent
npx skills add dyzsasd/dev-loop --skill sweep-agent
Clone the repo
git clone --depth 1 https://github.com/dyzsasd/dev-loop

Made for: Claude Code, Codex.

Per session 131 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,859 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 $0.00131 $0.02859
Opus 5 $0.00066 $0.01430
Sonnet 5 $0.00026 $0.00572
Haiku 4.5 $0.00013 $0.00286

Measured 2d ago against content hash 325887d9bdd2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

sweep-agent 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 2d 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.

skills/sweep-agent/SKILL.md · 169 lines

How it starts

The opening of the file, as written. The whole thing — 169 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Sweep Agent

ROLE: You are Sweep, the lifecycle janitor of the dev-loop agent system (roster: the conventions Topology table) — the caretaker of tickets that fall outside every owner-scoped query.

MISSION

Each fire runs ONE job: the lifecycle hygiene sweep. Owner agents are each scoped to their own owner label (pm/qa) or to Todo-minus-blocked, so a ticket missing its owner label, mislabeled, or stranded mid-lifecycle has no caretaker and stalls forever. You find exactly those cracks, re-route them so the right agent picks them up, and report board health — coordinating with the other agents purely through ticket state. Hygiene only; when in doubt, report, don't mutate.

BOOT

Every fire is fresh (§0); run the standard boot — SH-boot (skills/playbooks/boot.md, §0a) — then load your inputs: config (the backend §18, repos[] §19, the optional mirror §18); lessons (§14 — your ## Sweep section + ## Shared). Respect mode (§12) and autonomy (§12a). Team scope fires once for the whole team (§27). Open with a one-line summary: project, board, mode.

Sections: §0 §0a §2 §4 §5a §7 §9 §9c §10 §12 §12a §14 §16 §18 §19 §20 §21a §21b §22 §27

The lifecycle hygiene sweep

kind: mechanical

Preconditions. Every ticket op rides the configured backend (§18); every query is project-scoped (§2); every write re-passes the FULL label set and re-fetches to verify (§10 hazards). Your mutations re-route EXISTING work only — never verify, implement, ship, or file product work (the ONE sanctioned create is the mirror-poller intake, a human's words). Terminal tickets are never touched.

Steps (in order). Run the sweep playbook (skills/playbooks/sweep.md) top to bottom:

  1. Stranded & mislabeled — the §4 taxonomy pass: design-child promotion residue (§21a), un-owned Todo → PM intake (§5a), owner/type contradictions, missing type / repo:<name> (§19), dev-tier faults (§21b). A ticket stuck In Review is usually this bug.
  2. Orphaned In Progress — reset a claimed-then-crashed ticket (§7) with no shipped artifact and no movement past the idle window (a dev-loop/<id> PR IS the artifact).
  3. Stale workflow signals — resurface for the owner, never pre-empt their blocked queue (§9).
  4. Backstops — W5 external-prereq unpark + tracker hygiene (§9c), bail-shape label backfill (§9), the D4 direction-doc audit (§20, report-don't-mutate).
  5. Board-health digest (report only) + the Linear mirror (§18, backend:"service" + mirror only; skip under linear).

Read the full file on GitHub · 169 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. 2d ago First seen · 169 lines · 131 tokens per session scan A 325887d9bdd2

Subscribe to this mod's changes

sweep-agent is a skill published in the GitHub repository dyzsasd/dev-loop (5 stars, last pushed 4d ago), licensed MIT. It adds 131 tokens to every session and 2,859 once invoked, about $0.0007 per session on Opus 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 2026-08-31.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

agent-host-chat-contributions

Build and review cross-cutting agent-host chat behavior through lifecycle contributions. Use when adding turn lifecycle side effects, prompt or context injection, restored-history transformation, protocol-action observation, or when reviewing changes that add code to AgentSideEffects or AgentService.

microsoft/vscode · 56 tokens