code-generation
Run and troubleshoot repository code generation after detecting the generator contract.
Skill by matt-riley
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/matt-riley-code-generation
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: documents
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Shell scripts
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/code-generation
What's inside
- Skill name
code-generation- Description
- Run and troubleshoot repository code generation after detecting the generator contract. Use when schema, query, spec, or template inputs changed, or generated output may be stale.
- License
- GPL-3.0
- Metadata
- kind: task · owner: mattriley · version: 1.4.2 · audience: general-coding-agent · category: generation · maturity: stable
- Instructions
- 78 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Files (7)
- SKILL.md
- openai.yaml · agents/
- evals.json · evals/
- trigger-queries.json · evals/
- edge-cases.md · references/
- examples.md · references/
- check-generated.sh · scripts/
SKILL.md (first lines)
## Use this skill when - The task changes generator inputs such as SQL, proto, OpenAPI, templates, or generator config. - Build/test output points at stale generated packages, missing generated types, or outdated artifacts. - You need to verify whether generated output should be refreshed, checked in, or left to CI/runtime generation. ## Do not use this skill when - The request is to design a schema/spec/template format from scratch rather than run the repo's generator workflow. - The task is a generic explanation of code generation. - The failing area has no generator-backed inputs and no stale-generation symptoms. ## Inputs to gather - Which generators the repo uses and what inputs feed them. - The canonical generation command(s): wrapper target, script, or direct CLI. - Repo contract for outputs: - generated files committed to Git - generated locally for validation but ignored in Git - regenerated in CI/runtime from source inputs - Validation path: diff check, `check-generated` script, build/test target, or CI job. ## First move Inspect the repo's build docs, CI workflow, and `.gitignore` before regenerating or committing anything. Detect whether generated output is checked in, ignored, or only rebuilt in CI/runtime. ## Workflow 1. Identify the real generation entrypoints. - Prefer repo wrappers such as `make generate`, `task generate`, or checked-in scripts. - If wrappers are absent, locate the exact generator command for each tool (`sqlc generate`, `templ generate`, `buf generate`, `protoc`, etc.). 2. Regenerate only when it matches the repo contract. - Run generation after changing generator-backed inputs or when failures point to stale output. - If the repo uses multiple generators, run only the affected ones unless the wrapper target is the canonical path. 3. Branch on output handling. - **Committed output:** regenerate, inspect the diff, and keep the generated changes with the source changes. - **Ignored or CI-only output:** regenerate locally when needed for validation, but do not commit ignored artifacts unless the repo explicitly expects snapshots. - **Mixed repos:** follow per-generator rules; some outputs may be committed while others are ephemeral. 4. If tooling is missing or behaving differently, recover safely. - Install or invoke the pinned tool version the repo expects. - Check generator config, tool versions, and source inputs before blaming generated files. 5. Run the repo's follow-on checks. - Use the repo's diff or cleanliness check for generated files. - Build/test the affected area after generation succeeds. ## Outputs - The repo's canonical generator entrypoints and output-handling contract (committed, ignored, or mixed) identified for the touched inputs. - Regenerated artifacts for the affected generators only, with the resulting diff or clean status inspected against the repo contract. - Follow-on build, test, or generated-output checks confirming the regenerated state is valid. ## Guardrails - Never edit generated files by hand; change the generator inputs or config instead. - Do not assume `make generate` exists or that one command covers every generator in the repo. - Do not assume generated output belongs in Git; check `.gitignore`, CI, and repo docs first. - If no generator-backed inputs changed, do not force regeneration as the first move unless the failure clearly points to stale output. - Unexpected generated diffs usually mean changed inputs, config drift, or tool-version drift; inspect those before patching outputs. ## Validation - The relevant generation command completes successfully. - The diff or status matches the repo contract: - expected generated files changed and are ready to commit, or - ignored/ephemeral output was verified locally without being staged accidentally. - Follow-on build/test/check-generated steps pass. ## Reference files - Read `references/examples.md` when you need examples of stale-generation diagnosis or "regenerate before test" prompts. - Read `references/edge-cases.md` when the repo's committed-vs-ephemeral generation contract is unclear. - [`scripts/check-generated.sh`](scripts/check-generated.sh) — run with `--help` for a deterministic generation + diff check for committed-output workflows.
From matt-riley/agent-skills, GPL-3.0.
More from matt-riley/agent-skills
- 0.614score
agent-governance
matt-riley
Use this skill when building AI agents that call external tools, implementing policy-based access controls, adding semantic intent classification, creating…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
agent-instructions
matt-riley
Use when creating or updating agent instruction files (AGENTS.md for Pi, copilot-instructions.md for Copilot, per-path guides, or AGENTS.md router) —…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
agent-supply-chain
matt-riley
Use this skill when generating SHA-256 integrity manifests for agent plugins, verifying that installed plugins match their manifests, detecting tampered…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
agentic-eval
matt-riley
Use when designing or implementing an evaluation loop for AI agent outputs — reflection loops, evaluator-optimizer pipelines, LLM-as-judge scoring, or…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
api-smoke-validation
matt-riley
Use when running quick, repeatable API smoke tests after changes with hurl, including auth, status-code, response-shape, endpoint sanity, and pre-integration…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
aws-sdk-v2-to-v3-migration
matt-riley
Use when the codebase still imports from 'aws-sdk' v2, uses new AWS.*() constructors, .promise() calls, AWS.config.update(), DocumentClient, or s3.upload() —…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
circleci-to-github-actions-migration
matt-riley
Use when a repository still depends on .circleci/config.yml jobs, workflows, or artifacts that need to be replaced or phased out with GitHub Actions —…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
mocha-to-jest-migration
matt-riley
Use when a JavaScript or TypeScript suite still depends on Mocha, Chai, or Sinon and needs an incremental Jest migration.
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
project-references-migration
matt-riley
Use when a layered TypeScript workspace needs an incremental project-references migration without breaking package boundaries or editor resolution.
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
ast-grep
matt-riley
Use when you need ast-grep structural search, linting, or safe codemod rewrites, especially for requests that mention ast-grep, sg, codemod, rule files, or…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
ci-images
matt-riley
Validate CI checks locally and manage Docker image publishing workflows.
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
code-intelligence
matt-riley
Use when navigating or refactoring code — choosing between LSP, rg, or semantic search tools; an LSP call returns empty and you need to know if it failed; or…
SkillAFreshness: A★ 4Claude Code +9
Signals
- Stars
- 4
- Forks
- 0
- Open issues
- 1
- Last commit (this folder)
- Sep 29, 2026 (29 days ago)
- Last release
- Oct 23, 2026 (5 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- Version
- 1.4.2
- License
- GPL-3.0
- Repository
- github.com/matt-riley/agent-skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from matt-riley
- 0.614score
agent-instructions
matt-riley
Use when creating or updating agent instruction files (AGENTS.md for Pi, copilot-instructions.md for Copilot, per-path guides, or AGENTS.md router) —…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
agent-supply-chain
matt-riley
Use this skill when generating SHA-256 integrity manifests for agent plugins, verifying that installed plugins match their manifests, detecting tampered…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
agentic-eval
matt-riley
Use when designing or implementing an evaluation loop for AI agent outputs — reflection loops, evaluator-optimizer pipelines, LLM-as-judge scoring, or…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
api-smoke-validation
matt-riley
Use when running quick, repeatable API smoke tests after changes with hurl, including auth, status-code, response-shape, endpoint sanity, and pre-integration…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
aws-sdk-v2-to-v3-migration
matt-riley
Use when the codebase still imports from 'aws-sdk' v2, uses new AWS.*() constructors, .promise() calls, AWS.config.update(), DocumentClient, or s3.upload() —…
SkillAFreshness: A★ 4Claude Code +9 - 0.614score
agent-governance
matt-riley
Use this skill when building AI agents that call external tools, implementing policy-based access controls, adding semantic intent classification, creating…
SkillAFreshness: A★ 4Claude Code +9
Cite this page
code-generation skill. Agent Mart. https://agentmart.dev/skills/matt-riley-code-generation (last updated 2026-10-24)