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

/async-delegation

@7fcc555
Run it yourself

Use when you have several independent jobs for AI agents. Splits work into parallel tasks, hands them off like to coworkers, and reviews results in batches.

  • 2 files
  • 2.3 KB
  • Updated 2 months ago
  • GitHub
Use this Skill: https://skilld.dev/gh/erkamyaman/dhh-p-bloom-agent-skills/async-delegation

Nothing lands on disk. Nothing to clean up.

Fork this Skill

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

SKILL.md

≈44 tokens for metadata: the name and description. ≈344 when used: this file. ≈205 more on demand in 1 file.

Description uses 2.2% 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 ≈175 characters, or 2.2%.

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.

Async delegation

Inspired by DHH's Rails World 2026 keynote: https://www.youtube.com/watch?v=vDjW_dRyKXY

From the talk

DHH says the best way to work with agents is asynchronously: give a task like you would to a coworker, send the agent off, and review when something is ready, instead of sitting in a chat waiting for output. Splitting into parallel tasks, batch sizes and review order below are our additions, not from the talk.

Treat agents like coworkers: give a clear task, send them off, and review when something is ready. Don't sit watching tokens stream in a chat.

When to use

  • Three or more tasks that don't depend on each other.
  • A task that will take longer than a few minutes.

Steps

  1. List the tasks and mark which are independent. Only independent tasks run in parallel.
  2. Give each task its own brief (see outcome-prompting) and its own branch or workspace so changes don't collide.
  3. Start them all, then do something else.
  4. Review in one batch: read each result against its acceptance checks before merging anything.
  5. Send corrections as a short follow-up to the same task, not a fresh start.

Verification

Before merging, run the full test suite on the combined result, not only on each branch alone. Parallel changes can pass separately and still conflict.

See references/batching.md for how to size tasks and batches.

Source: SKILL.md on GitHub

No rule matched.

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 7fcc555. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub last month.

Activeupdated 2 months ago

README badge

README badge for erkamyaman/dhh-p-bloom-agent-skills/async-delegation