agentsclimarketplace
Sign in

Lets

Skill troels-im/agent-skills-playbook-template/.agents/skills/lets

Use when a user invokes /lets or asks to start, continue, or inspect a guided development flow such as create-product, create-prototype, continue-work, work-on-bug, improve-codebase, research-decision, or write-docs in an agent-skills-playbook-template repository.From its SKILL.md

Install
npx -y skills add troels-im/agent-skills-playbook-template --skill lets

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.5 KB, 517 tokens by cl100k_base, as published. Nobody here has run it

Lets

Use this skill as the low-token command surface for repository flows.

Workflow

  1. Read AGENTS.md.
  2. Read root index.md.
  3. Read flows/index.md.
  4. Select the smallest matching flow.
  5. Load the selected flow file.
  6. Create or update the flow run checklist named by that flow. Use UTC timestamped slugs for generated run names: run-YYYYMMDDTHHMMSSZ-short-slug.md.
  7. Select the tracker provider early. If the user did not specify GitHub, GitLab, or Linear, use local and create tickets in tickets/.
  8. Follow only the playbooks named by the flow and required by the current step.
  9. Prefix progress updates with the format required by the flow: XX% / YY% / ZZ%: message.
  10. Run just turn-test before every commit.
  11. At the final step of a flow, run just pr-test, fix failing PR tests, and repeat until the PR tests pass or a real blocker is recorded.

Prescribed Subagents

When a selected flow or playbook prescribes subagents, run it with subagents. If the host cannot run the prescribed subagent workflow, stop the loop immediately and follow playbooks/repository/prescribed-subagent-failure.md. Do not replace a prescribed subagent workflow with a sequential self-review unless the user explicitly approves that degraded fallback after the retrospective is queued.

User Input Gates

When a flow step needs user input, approval, a product decision, or feedback to continue correctly, pause the loop and ask. Do not infer missing decisions or continue on guessed requirements.

Before starting user-story inventory, specs, PRDs, acceptance criteria, or tickets, write assumptions to an active trace file using playbooks/planning/verify-assumptions.md and pause for user review.

Continue without asking only when the user explicitly says not to disturb them, not to ask, to proceed with best judgment, or gives an equivalent instruction. When doing so, record the assumption in the run checklist and keep it reversible where possible.

If no flow matches, say that clearly and route through playbooks/repository/choose-workflow.md.

Do not treat /goal or /loop as user commands for this template. Use /lets as the user-facing command and flows/ as the durable repository layer.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,834. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.