framework/core/skills/flowai-skill-plan-adr/SKILL.md
Record an architectural decision (chosen path + rejected alternatives + rationale) as a persistent MADR-style ADR in documents/adr/. Triggers — "record this decision", "write an ADR", "capture our rationale for picking X", "запиши ADR". Do NOT use for regular task planning (use flowai-skill-plan), already-decided trivial choices, or simple bug fixes.
npx skillsauth add korchasa/flow flowai-skill-plan-adrInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Record a non-trivial architectural decision in ./documents/adr/<YYYY-MM-DD>-<slug>.md so the rationale survives past the originating task file. ADR files are persistent (NOT gitignored) and capture: the context that forced the decision, the alternatives weighed, the chosen path, and the consequences. The format mirrors MADR (Markdown Architectural Decision Records).
This skill is invoked AFTER a decision has been made — or is being committed to in this conversation — not for "what should we do?" exploration. Use flowai-skill-plan for the latter. ADR captures the OUTCOME of decision-making.
</context>
<step_by_step>
Plan the work with a task management tool.
Identify the decision from the conversation or from a referenced task file:
Assign the ADR ID:
./documents/adr/ (if it exists) for *.md files. For each, read the frontmatter id: value. The new ID is ADR-{maxExistingNumber+1}, zero-padded to 4 digits. If the directory is empty or absent, the new ID is ADR-0001.Compute slug + filename:
./documents/adr/<YYYY-MM-DD>-<slug>.md where <YYYY-MM-DD> is today's date.Draft the ADR in chat:
id, status, date, optional tags array.## Context — 2–4 sentences on the situation and constraints.## Alternatives — bullet list. For each alternative:
- **<short name>** — 1-line description.
- Pros: <bullet list>
- Cons: <bullet list>
- Rejected because: <one sentence> (omit for the chosen alternative; mark it `(CHOSEN)`)
## Decision — 1–2 sentences stating exactly what was decided.## Consequences — bullet list of follow-on effects, both positive and negative. Mention any SDS/SRS updates implied (so the user knows to run further skills).Show draft to the user, ONE round of edits, then write.
Write the file:
mkdir -p documents/adr/ if missing.</step_by_step>
tools
Delegate a task to another AI IDE's CLI (codex / claude / opencode / cursor-agent) through an isolated-context subagent. Triggers on "delegate to <ide>", "have <ide> do <task>", "execute <task> in <ide>", "offload to <ide>". For one-shot relay or fan-out comparison use `ai-ide-runner` instead.
tools
Run prompts in Claude Code, OpenCode, Cursor, or Codex CLIs from the current session — pick one IDE, fan out across several, or compare models. You are a courier that relays the other runtime's stdout verbatim, do not synthesise your own answer. Use on "run in <ide>", "compare <ide> vs <ide>", "try on <model>", "which IDE handles X better", "run across models".
tools
Recommend which LLM model to use for a task. Use when asked "which model / best LLM for X", "pick a model for this task", or for a model shortlist ranked by live leaderboard evidence (coding, reasoning, agentic, tool-use, price, speed). Live-fetches public leaderboards and ranks models with per-axis rationale and citations.
development
Produce a comprehensive Product Requirements Document (PRD). Use when the user asks to write a PRD or formalize a feature's scope, goals, and success metrics.