Official AI Agent Plugins & Skills Marketplace for Antigravity (AGY), Claude Code, and Cursor
This repository also configures its own agents. See what agent-plugins tells them →
Agent Claude Code
Delegate to this subagent when you need changeset@2 artifacts produced from a git diff. Input is a git diff, optionally a linked spec@1, and optionally smith's criteriaevidence from the TDD run that produced this diff. A diff spanning multiple independent topics produces one changeset@2 per topic, never one bundled…
Agent Claude Code
Delegate to this subagent when the user has staged git changes and needs a conventional commit message written. Input is the staged diff. The agent reads shared/references/delivery/commits.md for type and scope conventions. Output is a JSON object containing commitmessage, type (feat, fix, docs, style, refactor, test…
Agent Claude Code
Delegate to this subagent when the user needs a pull request title and body written for a set of staged or committed changes. Input is a git diff and optionally a linked spec@1 or requirement@1. The agent reads the PR template from shared/references/delivery/github.md and writes from the reviewer's perspective — what…
Agent Claude Code
Delegate to this subagent when you need a release-artifact@2 generated from a collection of changeset@2 entries. Input is a list of changeset@2 JSON objects, the prior released version string, and a release date in YYYY-MM-DD format. Each input changeset already carries consumerimpact and semverimpact, set at…
Agent Claude Code
Delegate to this subagent before a PR is opened, to assemble the complete review package for the reviewer. Input is the changeset diff, optionally a linked spec@1 or requirement@1, test results, and any open questions the author wants answered. The agent bundles these into a structured review package — diff summary…
Agent Claude Code
Delegate to this subagent to audit a plugin directory for ecosystem conformance. Check manifests, declared files, agent structure, prompt-prefix identity, routing, contracts, instruction economy, runtime reference quality, and shared linkage. Use repository validators for deterministic checks. Treat length as a review…
Agent Claude Code
Delegate when creating a plugin or adding one agent to an existing plugin. Generate conformant manifests, skills, agents, routing, and shared linkage for the requested scope. Copy the constitution byte-for-byte. Return a JSON report of created files.
Agent Claude Code
Delegate to this subagent when the user wants to design a new JSON Schema for an inter-agent artifact — a structured payload passed between producers and consumers in the plugin ecosystem. Provide a description of what the artifact represents, who produces it, and who consumes it. The designer produces a JSON Schema…
Agent Claude Code
Delegate to this subagent when you have a completed requirement@1 (or free-text design intent) and need a formal component spec@1 drafted. Input is a requirement@1 and optionally design references (mockup description, existing component family, design-system tokens). Output is a spec@1 conforming to…
Agent Claude Code
Delegate to this subagent when a component spec@1 needs a definitive pass/fail judgment before entering the planning phase. Input is a spec@1 JSON object describing a UI component. This is an adversarial gatekeeper — the default disposition is fail, and the spec must earn a pass. The spec passes only if all four…
Agent Claude Code
Delegate to this subagent when a plan@1 needs adversarial review before implementation. Input is a plan@1 and its source spec@1. Check missing work, dependency order, prescription, batching, error handling, criterion coverage, and interface siblings against live evidence. Return per-issue repairs and a binding pass or…
Agent Claude Code
Delegate to this subagent when a plan@1 exists and you need a rough effort estimate before committing to execution. Input is a plan@1 JSON object. The agent produces a per-task time estimate in minutes, marks each task as parallelizable or sequential, lists blocking dependencies by task ID, and produces a totalminutes…
Agent Claude Code
Delegate to this subagent when you have a completed spec@1 and need a sequenced, executable implementation plan. Input is a spec@1 (or its specfilepath, read from disk if provided). Output is a plan@1 conforming to shared/schemas/[email protected], with specfilepath and spechash propagated from the spec. Decomposes into…
Agent Claude Code
Invoke once per candidate after scanner and, for T7 or T10, boundary tracing. Try to refute reachability from current code and return candidate-assessment@1.
Agent Claude Code
Trace T7 and T10 candidates across execution boundaries. Supports Rust, TypeScript, JavaScript, Python, and Go. Return field-survival-map@1.
Agent Claude Code
Invoke before scanning. Detect affected languages, trace reachable files from live entry points, and inventory project capabilities and current workspace state. Return workspace-manifest@1; do not judge defects.
Agent Claude Code
Invoke after recon. Load the hazard pack for each scoped language, scan only live files, and emit every matching candidate@1 without verdicting. Supports Rust, TypeScript/JavaScript, Python, and Go.
Agent Claude Code
Delegate when finding-report@2, implementation-review@2, or implementation-result@1 identifies a structural defect class. Inspect the live semantic model and persisted architecture, then produce a spec@1 that makes recurrence impossible or machine-detectable.
Agent Claude Code
On-demand diagnostic: given a specfilepath and a workspace root, verifies the current code still implements every acceptance criterion in the spec. Reads the spec from disk, not from context, and reads implementation/tests from the workspace. When a prior changeset@2's criteriaevidence is available, uses its pointers…
Agent Claude Code
Delegate to this subagent when you need to verify that a draft spec@1 is grounded in its source artifacts — the requirement@1 and optional research-report@1 — before it proceeds to auditing or gating. Input is a spec@1 draft plus the originating requirement@1 and any research-report@1 used to produce it. For each…
Agent Claude Code
Delegate to execute one plan task. Read persisted criteria, design and implement the change, prove it with tests, and record exact evidence. Commit only when the caller supplies commitauthorized: true; otherwise return a null commit SHA. Report spec contradictions or architecture needs instead of forcing a pass. Scope…
Agent Claude Code
Delegate after implementation. Use available project-native mutation tooling against changed behavior, or a controlled deliberate fault when appropriate. Analyze survivors and return mutation-report@2; unavailable tooling becomes a coverage gap.
Agent Claude Code
Delegate after an implementation batch and discrimination check. Review scope, verification fit, related defects, semantic models, and architecture. Return implementation-review@2; route structural families to scribe:architect.
Agent Claude Code
Delegate to this subagent after recon produces a source map. Input is the source map from recon plus the original research question. The agent reads every identified source and extracts findings relevant to the question, with a citation, confidence level (confirmed, likely, or assumed), and source attribution for…
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: