Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Sweep · fallow-rs | DecimalAI

agent toolingby fallow-rs

Sweep

fallow-rs/sweep

Source fallow-rs/fallow ↗License MIT

Audit the current Fallow session for missed work, incomplete verification, stale documentation, companion drift, or cleanup before final completion.

NewMaintained

Automatic

Coding pack

Last verified 9/28/2026

What it changes

Passed 6 of 12 test cases with the skill, and 7 of 12 without it.

Without skill57%

With skill48%

Checks it passed that the model alone missed

Maps active plan against live pull-request state
Requires removing generated artifacts from workspace
Validates documentation updates for external interfaces

and 2 more in the full report

4 fixed7 passed either way6 still failing6 broke

Pass-rate lift: -9 pp
Tokens: +7%
Agent turns: 0%
Response time: +10%

Model gemini-3.6-flashMeasured 9/28/2026v2 · latest is v411 of 23 cases passed with skill · 12 comparable

Tokens, turns and time describe resource use, not a dollar price.

This benchmark measured an earlier version of the skill.

A newer attempt on gemini-3.6-flash on 10/4/2026 did not produce a publishable comparison (not comparable). The snapshot above is the previously measured run.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/fallow-rs-sweep/SKILL.md https://app.decimal.ai/s/fallow-rs-sweep/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add fallow-rs-sweep, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/fallow-rs-sweep/ and pins it in .decimal/skills.lock

decimalai skills add fallow-rs-sweep

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/fallow-rs-sweep/SKILL.md https://app.decimal.ai/s/fallow-rs-sweep/SKILL.md

Raw file: /s/fallow-rs-sweep/SKILL.md ↗ · machine-readable summary at /s/fallow-rs-sweep.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 4

Prompt

The integration test suite passed yesterday morning before we merged three upstream changes to main on the order-service repo. Is yesterday's green build sufficient for merging?

What was checked, without and with the skill
Checked	Without skill	With skill
Requires re-running stale verification against current branch state	Passed	Passed
Maps active plan against live pull-request state	Missed	Passed

case-07 · gemini-3.6-flash · 9/28/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md25 linessingle file

Sweep

In this repository, use this sweep instead of a general sweep workflow.

Re-read the user's full request, active plan, diff, and live pull-request state.
Map every acceptance criterion to authoritative evidence.
Check for missed consumers, output formats, filters, docs, skills, schemas, CI scripts, companion repositories, and private/public boundaries.
Re-run any weak or stale verification.
Fix in-scope omissions and document genuine separate follow-ups.
Verify branches, worktrees, temporary worktrees, and generated artifacts are clean.
List the lessons of the task: corrections, failures that cost time, and skipped steps. Keep a lesson only when all four filters pass:
- it stays true after paths, versions, and counts change;
- it changes a future decision, not only the amount of text to read;
- no skill or doc covers it yet (sharpen a weak rule, do not copy it);
- it belongs to a skill, doc, or surface that this task used.
Store each kept lesson in the strongest form: a type or schema, then a lint, test, or CI check, then a script. Write text only when none fits.

Finding no obvious bug is not completion. Prove every requested outcome.

Track record

Benchmark

6 / 12

test cases passed, vs 7 / 12 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 9/30/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

fallow-rs / fallow

.agents/skills/sweep/SKILL.md

View on GitHub ↗

License MITImported 9/30/2026Version v4

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Sweep

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills