Index of every page on this site: llms.txt
All skills
sandharbor avatar

/agent-task

@62fef7a
Run it yourself

List and run agent tasks defined in _module/agent_tasks/ directories across the repo

  • 1 file
  • 1.5 KB
  • Updated 2 months ago
  • GitHub
Use this Skill: https://skilld.dev/gh/sandharbor/meadow/agent-task

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈24 tokens for metadata: the name and description. ≈345 when used: this file.

Description uses 1.2% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈97 characters, or 1.2%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

/agent-task — Run Agent Tasks

Run agent tasks defined in _module/agent_tasks/ directories throughout the repo using the tools/agent_task/run_agent_task runner.

When invoked with no arguments (/agent-task)

Scan for all agent task markdown files and list them:

find . -path '*/_module/agent_tasks/*.md' -not -path '*/node_modules/*' | sort

Display the results as a list showing the full paths so the user can see where each task lives.

When invoked with arguments (/agent-task <name> [options])

  1. Resolve the task file — match the user's <name> against the discovered agent task files. Use fuzzy/substring matching (e.g. "nodespec" matches nodespec_chaos_check.md). If the match is ambiguous, ask the user to clarify.

  2. Determine run count — look for a number in the user's request (e.g. "run nodespec chaos 30 times", or "-n 30"). Default to 1 if not specified.

  3. Run it — execute the runner, which invokes claude --print in a loop:

    ./tools/agent_task/run_agent_task -n <count> -v <resolved_task_file>

    Run this in the background since agent tasks are long-running. The runner logs to tools/agent_task/run_agent_task.log — tell the user they can tail -f that file to watch progress.

    Use the -v (verbose) flag by default so tool calls are visible in the log.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at 62fef7a. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 weeks ago.

Activeupdated 2 months ago

README badge

README badge for sandharbor/meadow/agent-task