/octocode-rfc-generator
Use when a consequential change needs a decision before coding: write or improve an RFC, design doc, architecture proposal, migration plan, option comparison, rollout plan, or measurable implementation contract.
$ npx -y skills add bgauryy/octocode-mcp --skill octocode-rfc-generator --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/octocode-rfc-generator
Context preview
The summary Claude sees to decide when to auto-load this skill.
Use when a consequential change needs a decision before coding: write or improve an RFC, design doc, architecture proposal, migration plan, option comparison, rollout plan, or measurable implementation contract.
SKILL.md
octocode-rfc-generator.SKILL.mdname: octocode-rfc-generator
description: "Use when a consequential change needs a decision before coding: write or improve an RFC, design doc, architecture proposal, migration plan, option comparison, rollout plan, or measurable implementation contract."
Octocode RFC Generator
Produce evidence-backed decisions that an implementer and reviewer can execute. Flow: `UNDERSTAND → RESEARCH → PREREQUISITES? → COMPARE → WRITE → CLOSE QUESTIONS → KPIs → VALIDATE → DELIVER`. For existing RFCs, run `REASSESS` (see `references/workflow.md` § Reassess existing RFCs) instead of WRITE — audit against live code, not prior checkboxes.
Lobby rules
- Skip RFC mode for trivial edits. Ask one focused question when uncertainty changes shape, owner, scope, or decision criteria.
- Compare at least two alternatives, including do-nothing, unless the user explicitly requests one implementation plan.
- Recommendations require verifiable facts; cite exact anchors and commands/checks that actually ran.
- `RFC.md` owns goals, scope, and decision. Other files link to its anchors rather than restating them.
- Resolve every open question with evidence or an explicit deferral; no recommendation may rest on an uncertain claim.
- Order implementation by dependency, not estimates; bind requirements to acceptance and verification.
- Reviewing, rating, or cleaning up `.octocode/rfc/` (delete/implemented/fixed calls) always requires a dated `## Audit Reasoning` block written into the RFC itself, backed by fresh reads of the live code.
- Never assert RFC status from memory or from another RFC's claims.
Artifact route
- Small, reversible, single-package work: produce only `RFC.md` with plan, acceptance, and inline references.
- Otherwise, after save approval, create `<workspace>/.octocode/rfc/{name}/`: `RFC.md` (decide), `PREREQUISITES.md` (ready, existing code only), `IMPLEMENTATION.md` (build), `KPI.md` (verify), and `RESOURCES.md` (source appendix).
Smart routes — load only what the current step needs
- Before drafting, load `references/workflow.md` — select mode, gates, claim ledger, artifact set, traceability, validation, and delivery.
- When gathering evidence, load `references/octocode.md` then `references/research-playbook.md` — delegate exact research and keep claims auditable.
- When writing the decision, load `references/rfc-template.md` — structure options, goals/non-goals, reversibility, and pre-mortem.
- When existing code has readiness work, load `references/rfc-prerequisites.md` before planning — expose baselines, blockers, owners, and setup.
- When building the execution plan, load `references/rfc-implementation.md` — close questions, order dependencies, and define rollout/rollback.
- When defining acceptance, load `references/rfc-kpi.md` — connect user stories, metrics, decision rules, and verification in a traceability matrix.
- When preserving sources, load `references/rfc-resources.md` — record provenance without moving decisive citations out of the RFC.
- When improving this skill, prefer `octocode-graph-eval`; otherwise load `references/improve-loop.md` — enforce measurable accept/revert.
Related routes and verification
- Use `octocode-brainstorming` before RFC when worth-building is unresolved; `octocode-research` to close factual questions; `octocode-graph-eval` for KPI rigor.
- Use `octocode-awareness` while coordinating drafts; `octocode-skills` when changing this skill folder.
- Before delivery run `scripts/eval-rfc.mjs --case <id>` — validate the document contract and report the real result.
Read more
name: octocode-rfc-generator description: "Use when a consequential change needs a decision before coding: write or improve an RFC, design doc, architecture proposal, migration plan, option comparison, rollout plan, or measurable implementation contract."
Octocode RFC Generator
Produce evidence-backed decisions that an implementer and reviewer can execute. Flow: `UNDERSTAND → RESEARCH → PREREQUISITES? → COMPARE → WRITE → CLOSE QUESTIONS → KPIs → VALIDATE → DELIVER`. For existing RFCs, run `REASSESS` (see `references/workflow.md` § Reassess existing RFCs) instead of WRITE — audit against live code, not prior checkboxes.
Lobby rules
- Skip RFC mode for trivial edits. Ask one focused question when uncertainty changes shape, owner, scope, or decision criteria.
- Compare at least two alternatives, including do-nothing, unless the user explicitly requests one implementation plan.
- Recommendations require verifiable facts; cite exact anchors and commands/checks that actually ran.
- `RFC.md` owns goals, scope, and decision. Other files link to its anchors rather than restating them.
- Resolve every open question with evidence or an explicit deferral; no recommendation may rest on an uncertain claim.
- Order implementation by dependency, not estimates; bind requirements to acceptance and verification.
- Reviewing, rating, or cleaning up `.octocode/rfc/` (delete/implemented/fixed calls) always requires a dated `## Audit Reasoning` block written into the RFC itself, backed by fresh reads of the live code.
- Never assert RFC status from memory or from another RFC's claims.
Artifact route
- Small, reversible, single-package work: produce only `RFC.md` with plan, acceptance, and inline references.
- Otherwise, after save approval, create `<workspace>/.octocode/rfc/{name}/`: `RFC.md` (decide), `PREREQUISITES.md` (ready, existing code only), `IMPLEMENTATION.md` (build), `KPI.md` (verify), and `RESOURCES.md` (source appendix).
Smart routes — load only what the current step needs
- Before drafting, load `references/workflow.md` — select mode, gates, claim ledger, artifact set, traceability, validation, and delivery.
- When gathering evidence, load `references/octocode.md` then `references/research-playbook.md` — delegate exact research and keep claims auditable.
- When writing the decision, load `references/rfc-template.md` — structure options, goals/non-goals, reversibility, and pre-mortem.
- When existing code has readiness work, load `references/rfc-prerequisites.md` before planning — expose baselines, blockers, owners, and setup.
- When building the execution plan, load `references/rfc-implementation.md` — close questions, order dependencies, and define rollout/rollback.
- When defining acceptance, load `references/rfc-kpi.md` — connect user stories, metrics, decision rules, and verification in a traceability matrix.
- When preserving sources, load `references/rfc-resources.md` — record provenance without moving decisive citations out of the RFC.
- When improving this skill, prefer `octocode-graph-eval`; otherwise load `references/improve-loop.md` — enforce measurable accept/revert.
Related routes and verification
- Use `octocode-brainstorming` before RFC when worth-building is unresolved; `octocode-research` to close factual questions; `octocode-graph-eval` for KPI rigor.
- Use `octocode-awareness` while coordinating drafts; `octocode-skills` when changing this skill folder.
- Before delivery run `scripts/eval-rfc.mjs --case <id>` — validate the document contract and report the real result.
Evidence-first code research for AI agents and developers. Octocode researches your local code and external code alike (GitHub repos, PRs, npm) with one toolset: ripgrep + AST search, trees, precise reads, and LSP.
Repo: bgauryy/octocode-mcp
Other skills on octocode.
- /octocode-benchmark
Use when planning, running, grading, or reporting the by-hand Octocode research benchmark — pairwise matchups (Octocode anchor vs one baseline: gh+RTK, gh+Headroom, or plain gh) over markdown questions, with a fresh isolated runner agent per (question, arm, pass), one blind
Open skill - /octocode-awareness
Use when coordinating work in a shared repo — multi-agent or solo across sessions: collision avoidance, handoffs, verification debt, durable memory/wiki, hooks setup/debug, and repo learning before planning, editing, reviewing, or testing.
Open skill - /octocode-brainstorming
Use when an idea needs disciplined exploration before building: generate options, test worth-building, map adjacent solutions, challenge assumptions, narrow scope, or choose Build RFC / Prototype / Narrow / Park — even without saying brainstorm.
Open skill - /octocode-chrome-devtools
Use when browser debugging, scraping a known URL efficiently, or running a multi-step page workflow needs real Chrome DevTools evidence: network, console, performance, DOM/CSS, screenshots/PDF, security, cookies/storage, click/fill/search flows, or auth-gated live pages via CDP
Open skill - /octocode-documentation
Use when writing or updating docs: README, API docs, runbooks, AGENTS.md, CLAUDE.md, ADRs, Diátaxis restructuring, or codebase documentation for humans or coding agents.
Open skill - /octocode-graph-eval
Use when measuring whether a change helped or looping until a target metric is hit — for one agent loop or a graph of loops (multi-agent workflow): ACCEPT/REVERT, keep/discard, goal→KPI contracts, eval suites, graders, held-out checks, benchmarks, TDD failing-case-first,
Open skill

