feO2x/guided-coding / skills/guided-coding-setup
guided-coding-setup
VerifiedSet up or upgrade Guided Coding repository instructions, plan storage, and documented feedback-loop commands. Run only when explicitly requested by the user.
Set Up Guided Coding
Your goal is to set up or upgrade Guided Coding in the current repository. After finishing, report what you changed, give the customization tips below, and stop.
Guided Coding needs these artifacts:
AGENTS.mdat the repository root, listing the feedback loops and pointing toai-plans/AGENTS.md.ai-plans/, the folder holding all plans and Plan Deviations documents.ai-plans/AGENTS.md, describing the folder, its file naming rules, and how Frozen Plans are treated.
Create missing files and folders, update outdated Guided Coding instructions, and leave current content unchanged. Preserve unrelated content, including project-specific instructions and user-authored notes.
1. Inspect the Repository
Read the root AGENTS.md and ai-plans/AGENTS.md if they exist, plus build manifests, task runners, scripts, and CI configuration.
Identify the feedback loops and their exact commands. Look for:
- compilers, type checkers, static analyzers, and linters
- automated tests, code coverage, and mutation testing tools
- benchmarks
- dependency, secret, container, and source-code security scans
Only select commands that are immediately executable in the repository. Do not list tools merely because they are common for the detected language.
2. Ensure the Root AGENTS.md
Create AGENTS.md in the repository root if it does not exist. Otherwise, make the smallest update that adds what is missing and corrects what is outdated. Match existing sections by meaning, not by exact heading.
Ensure it contains:
## Feedback Loops: each command and what it verifies. Report to the user when no feedback loops could be found, and warn thatguided-coding-write-planrefuses to write plans until at least one is listed.## Guided Coding: a link toai-plans/AGENTS.md, noting that it holds the file naming conventions and the rules for working with Frozen Plans. Link to those rules instead of repeating them in the root file.
If both sections already exist and are current, leave the file alone.
3. Ensure ai-plans/AGENTS.md
Create ai-plans/ if it is missing. Compare ai-plans/AGENTS.md with assets/ai-plans-AGENTS.md relative to this skill file.
- If the file is missing, copy the asset.
- If the file exists with an older
guided-coding-versionmarker or without one, update the Guided Coding paragraphs and the marker. Keep every repository-specific addition, such as a custom plan structure, naming conventions, or notes about legacy filenames.
4. Report and Give Customization Tips
List the files you created, updated, or left unchanged, the feedback loops you documented, and which commands you ran. Ask the user to verify the changes.
Then explain how Guided Coding can be customized, so the user knows the defaults are only a starting point:
-
Plan structure:
guided-coding-write-planuses a default structure with## Rationale,## Acceptance Criteria, and## Technical Details. It also ships the RDAT structure (rdat-plan-structure.md), which adds a## Decisionssection with numbered Decisions as a summary for the reviewer. Name a structure file when invokingguided-coding-write-plan, or name it or describe your own structure inai-plans/AGENTS.mdto override the default. -
Filenames: the skills use the placeholders
<timestamp>(UTC,YYYY-MM-DD-HHMM),<ticket-id>, and<short-title>(kebab-case, shortened ticket title). Conventions for these, e.g., a<ticket-id>format or a maximum length for<short-title>, belong inai-plans/AGENTS.md. -
Feedback loops: plans should reference the feedback loops in the root
AGENTS.mdto make acceptance criteria verifiable. Adding or removing a feedback loop later only requires updating that list. -
Plan review: configure
guided-coding-plan-and-reviewin an optional## Plan Reviewsection inai-plans/AGENTS.md. It supports a reviewer list with a model preference for each, a round limit (default 3), and a policy for findings that challenge the user's explicit decisions.collect(default) reports them as open for the user.pausestops the round and asks the user.Each round starts one subagent per reviewer. The maximum number of subagent runs is the number of reviewers multiplied by the round limit, plus one replacement for each reviewer that fails.
-
Deviations: the threshold for what counts as a material deviation can be tightened or relaxed in
ai-plans/AGENTS.md.
Repository-specific conventions in AGENTS.md files take precedence over the skills' defaults.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |