Copy. Paste. Run. — Production-tested Agent · Skill · Squad templates for Multica, bilingual (Chinese/English)
Skill Claude CodeCodex
Land API contract artifacts to the API collaboration platform (default Apifox). Used by @BackendDev to upload API contract for downstream frontend / test consumption. Swappable platform.
Skill Claude CodeCodex
CI/CD artifact-orchestration skill (placeholder shell): after G2 PASS and code push, call the underlying CI/CD platform skill to trigger dev/sit builds, write back to the Issue, and return the deploy URL. Parameters auto-discovered from the CI API; the orchestration layer never hardcodes parameter names. Concrete…
Skill Claude CodeCodex
Land technical design artifacts to the code repo or knowledge platform (default Git repo / Confluence). Used by @Architect to upload technical design for downstream implementation and test consumption. Swappable platform.
Skill Claude CodeCodex
Land product requirement / PRD artifacts to the requirement knowledge platform (default Confluence). Used by @ProductManager to upload PRD and return a link for downstream design / dev / test consumption. Swappable platform.
Skill Claude CodeCodex
Land test case / report artifacts to the case-management platform (default local XMind → Jira). Used by @Tester to upload cases and return a link for downstream acceptance consumption. Swappable platform.
Skill Claude CodeCodex
Land UI / interaction design artifacts to the design collaboration platform (default Figma). Used by @Designer to upload visuals and return a stable link for downstream frontend consumption. Swappable platform.
Skill Claude CodeCodex
Integrate CI hard gates into the target repository and make gatekeeping read the CI verdict. Used for deploying gates, reading check-runs for G2, and falling back to soft gates when there's no CI.
Skill Claude CodeCodex
Implement an approved technical plan with minimal unnecessary changes. Used for coding, adding tests, and running verification.
Skill Claude CodeCodex
Platform-layer skill (placeholder shell): read/write capability for Confluence / Wiki-class systems — pull existing pages for Agents to consume, land PRD / design artifacts and return a stable page link. Decoupled from the Issue system; called by artifact-orchestration skills. Concrete platform URLs / spaces / parent…
Skill Claude CodeCodex
Platform-layer skill (placeholder shell): read/write capability for CI/CD systems (Jenkins-class) — trigger parameterized builds, poll status, fetch console logs. Parameters are auto-discovered; never hardcode. Credentials injected via runtime env. Concrete CI URL / Job list in config.yaml, never in role prompts.
Skill Claude CodeCodex
Platform-layer skill (placeholder shell): read/write capability for Issue-tracking systems (JIRA-class) — Issue query, upstream-Wiki-link parsing, Story creation, transition, scheduling, description write-back, notification. Decoupled from the Wiki platform; called by artifact-orchestration skills. Concrete URLs /…
Skill Claude CodeCodex
Turn an Issue into a clear, actionable engineering task. Used for requirement clarification, scope confirmation, and acceptance-criteria definition.
Skill Claude CodeCodex
Architecture-design dedicated review framework. Called by ArchReviewer to professionally analyze Architect's technical design (soundness/extensibility/acceptance alignment/tech risk), output PASS/FAIL + fix list, report to Leader.
Skill Claude CodeCodex
Backend-implementation dedicated review framework. Called by BackendReviewer to professionally analyze BackendDev's API contract & implementation (contract quality/design fit/error handling/unit-test sufficiency/acceptance mapping), output PASS/FAIL + fix list, report to Leader.
Skill Claude CodeCodex
UI-design dedicated review framework. Called by DesignReviewer to professionally analyze Designer's UI/interaction design (interaction soundness/accessibility/design-system consistency/edge states), output PASS/FAIL + fix list, report to Leader.
Skill Claude CodeCodex
Frontend-implementation dedicated review framework. Called by FrontendReviewer to professionally analyze FrontendDev's output (UI/API-contract fit/component quality/unit-test sufficiency/acceptance mapping), output PASS/FAIL + fix list, report to Leader.
Skill Claude CodeCodex
Requirements/PRD dedicated review framework. Called by ProductReviewer to professionally analyze ProductManager's PRD (scope clear/goal explicit/acceptance testable/constraints complete/open items), output PASS/FAIL + fix list, report to Leader.
Skill Claude CodeCodex
Test-artifact dedicated review framework. Called by TestReviewer to professionally analyze Tester's cases/report/coverage doc (coverage depth/acceptance mapping/coverage reasonableness/conclusion truth/contract consistency), output PASS/FAIL + fix list, report to Leader.
Skill Claude CodeCodex
Produce a minimal technical plan based on the existing code. Used for architecture analysis, impact assessment, and implementation-plan design.
Skill Claude CodeCodex
Test-automation skill (placeholder shell): Tester phase-3 automation — after G2.5 the deploy environment is ready, run the team's automation tool against the scenario cases and produce G3 input. Concrete tool (Apifox / Postman / Playwright / pytest / etc.) onboarded by the team; credentials injected via runtime env…
Skill Claude CodeCodex
Test design function: produce feature cases and API cases from the requirement / design / API contract, execute automation-first, and deliver a test report mapped item by item to the acceptance criteria. Used for feature cases / API test cases / test execution / test reports.
Skill Claude CodeCodex
Gatekeeping function: objectively check whether an artifact satisfies the acceptance criteria. Triggered by the Leader to rerun at gate points (G1/G2/G3); members can also use it for self-check. Used for design review / implementation acceptance / test-report review.
Skill Claude CodeCodex
A Python client for the self-hosted Multica platform, which coordinates AI agents. It retrieves workspaces, agents, and recent agent activity through the platform API.
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: