Agent Claude Code
Add a new AI CLI adapter to agnostic-ai end to end. Use when a new target tool is added.
Agent Claude Code
Add a new AI CLI adapter to agnostic-ai end to end. Use when a new target tool is added.
Agent Claude Code
Close a confirmed target-audit finding end to end and open a PR. Never merges.
Agent Claude Code
Curate CHANGELOG.md Unreleased notes to the layout and entry rules. Use after merges and before a release.
Agent Claude Code
Review a Go diff in agnostic-ai for bugs, style, and cross-adapter issues. Use before merging a PR.
Agent Claude Code
Triage sync --check drift between source specs and emitted per-CLI configs. Use when sync --check fails.
Agent Claude Code
Audit one batch of agnostic-ai targets against current vendor docs, reporting evidence-backed drift and meaningful project capability changes.
Skill Claude CodeCodex
Cut a new agnostic-ai release end to end. Use when the user wants to ship a new version (tag + GitHub Release).
Skill Claude CodeCodex
Rebuild and serve the WASM playground at docs/playground/. Use after editing the playground source or any code reachable from cmd/agnostic-ai-wasm.
Skill Claude CodeCodex
Regenerate docs/schemas/config.schema.json from the Config struct. Use after editing internal/config/config.go.
Skill Claude CodeCodex
Run agnostic-ai sync --check, interpret the diff, and pick the right next step. Use after editing specs or adapters.
Skill Claude CodeCodex
Author a new agnostic-ai spec (agent, skill, rule, or hook). Use when the user wants to add a new prompt, convention, or hook to their project.
Skill Claude Code
Audit every agnostic-ai target against current vendor docs, report evidence-backed drift, and report project capability changes. Use for the recurring adapter truth check or when a tool ships a new config feature.
Plugin Claude Code
Bundles 4 skills · 206 tokens together
One spec, every AI CLI. Install the CLI, scaffold specs, import an existing config, and keep every tool's generated files in sync.
Agent Claude Code
Set up or reshape project root agent instructions from current source and official model guidance.
Agent Claude Code
Audit project agent context for duplication, stale guidance, unsafe instructions, and missing boundaries.
Skill Claude CodeCodex
Set up or review project agent instructions from an evidence-based context inventory. Use when adding or trimming files under .agnostic-ai/.
Skill Claude CodeCodex
Curate this tool's auto memory. Use when the memory index is near its load limit or the stored notes look duplicated or stale.
Skill Claude Code
Review every open PR with the code-reviewer agent and an adversarial review, apply the findings, and merge each one once CI is green. Use when asked to ship, sweep, or merge open PRs.
Skill Claude CodeCodex
Decide whether a change to the spec format needs a migration, and write it so agnostic-ai migrate rewrites old specs safely. Use when a change renames, replaces, deprecates, removes, or tightens any field users write under .agnostic-ai/ or in agnostic-ai.yaml.
Skill Claude Code
Take one GitHub issue from branch to merged PR with TDD. Use when asked to work, fix, or implement an issue by number.
Skill Claude Code
Work the open issues that are unassigned or assigned to you, one at a time through agnostic-ai-gh-issue. Use when asked to work the backlog.
Skill Claude CodeCodex
Write or resume a session handoff between AI tools.
Skill Claude CodeCodex
Recall or curate the project memory that every AI coding tool shares.
Skill Claude CodeCodex
Terse caveman voice: answer first, fluff gone, every technical fact kept. Use for /caveman, "caveman mode", "talk like caveman", "be brief", "less tokens". Stays on until "stop caveman" or "normal mode".
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: