Skip to content
Skills Library
Programming · Automation

until

/until

Keep a coding agent working until a check such as the test suite passes, with Trailboss's until loop, and steer running loops. Use when asked to have an agent keep going until tests, a build or a linter pass.

Licence
MIT
Last repository change
13.11.2026

Assessment and security

96/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content31 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 13.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Keep a coding agent working until a check such as the test suite passes, with Trailboss's until loop, and steer running loops. Use when asked to have an agent keep going until tests, a build or a linter pass.

Original author description

Keep a coding agent working until a check such as the test suite passes, with Trailboss's until loop, and steer running loops. Use when asked to have an agent keep going until tests, a build or a linter pass.

How it works

SKILL.md has no level-two headings · 18 lines, 214 words

SKILL.md preview
---
name: until
description: Keep a coding agent working until a check such as the test suite passes, with Trailboss's until loop, and steer running loops. Use when asked to have an agent keep going until tests, a build or a linter pass.
requires: trailboss >= 0.4, api 1
---

# Keep an agent going until a check passes

First run `trailboss version --json`. If it fails, or `version` is below 0.4 or `api` isn't 1, stop and tell the user to update Trailboss.

1. Pick the session (`AGENT:ID`, a session id, or its terminal from `trailboss --json live`) and a check the user agrees to. It runs as the user, without a shell, in the session's folder after every turn: `-- cargo test`, or `-- sh -c 'npm run build && npm test'`.
2. Set limits that fit the job, flags before `--`:

   ```sh
   trailboss until --max-rounds 5 --max-cost 2 --max-hours 1 --detach t3 -- cargo test
   ```

   Defaults: 10 rounds, $5, 2 hours, and a 10-minute `--check-timeout`. `--detach` keeps it running after this terminal closes; without it, the command holds this terminal until the loop stops.
3. Watch: `trailboss --json until ls` (each loop's `state`, `round`, `last_passed`, `cost_usd`, and `log` for a detached one).
4. Steer: `trailboss until pause|resume|stop <loop id|session|terminal>`. Use `all` only when the user says so.

When it stops, report why (`stopped` and `reason` in the last `--json` line or the log): passed, a limit, the same failure with no changes, no progress, or the session ended. Don't start a second loop on the same session; Trailboss refuses it.

Versions

RepositoryScoreLicence
This page · trailboss-ai/trailboss-skills96MIT
alesaiani/agentflow95Apache-2.0

Database snapshot: 13.11.2026 · package version 94c922cd3114 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 92fix-linter-test
    Debug and fix a failing MegaLinter linter test. Use when a linter test fails in CI or locally.
  • 100run-tests
    Build Amethystora from the last commit, boot it in the VM lab, run the test suites, and fix what fails until it passes. Use only when the user explicitly asks to run the tests ("run the tests", "build and test", "test it in the VM").
  • 86lifecycle-loop
    Turn the request into a goal card, run its checks, and loop test-first work and build-and-test until every check passes or progress stalls. Use at the start of a task, whenever a check fails, and when a goal tool or a goal-verification prompt asks whether you are done.
  • 90automate-tests
    Turn test cases into runnable automated tests, drive a real browser to explore and verify behavior, run the tests, and auto-fix them until they pass. The execute step of the `qa` skill; use it directly for a single test-writing or test-fixing task the user scopes themselves…
  • 100grinding-until-pass
    Keep iterating on code changes until the tests pass, the build succeeds, or linting is clean. Runs in a tight loop of fix → run → check → repeat. Use when you want the agent to autonomously grind through test failures or build errors.
  • 92update-linter-version
    Update a linter's pinned version in its descriptor YAML. Use when upgrading a linter tool to a new release.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.