Agent Claude Code
Independent code-quality and security reviewer. Use when changes are ready for review or when an adversarial second pass is needed.
not rated 18
changed 27d ago A 28 tokens
Agent Claude Code
Scoped implementation specialist. Use for an independent, well-defined plan phase with explicit file ownership.
not rated 18
changed 27d ago A 23 tokens
planner
03Agent Claude Code
Implementation-planning specialist. Use when a feature, migration, or multi-file change needs a decision-complete plan before coding.
not rated 18
changed 27d ago A 27 tokens
researcher
04Agent Claude Code
Research specialist. Use proactively when a task needs current technical facts, option comparison, or source-backed recommendations.
not rated 18
changed 27d ago A 24 tokens
tester
05Agent Claude Code
Test-validation specialist. Use after implementation to select, run, and assess the smallest relevant test suite.
not rated 18
changed 27d ago A 22 tokens
PreToolUse
06Hook Claude Code
One of 3 in hooks.json
Runs before the agent uses a tool for Write tool calls, executing privacy-block.mjs, scout-block.mjs and descriptive-name.mjs via node with --platform (3 commands). From Unibean9/harness-skills.
not rated 18
changed 27d ago A
tokens not measured
Hook Claude Code
One of 3 in hooks.json
Runs when you submit a prompt, before the agent sees it, executing language-prompt.mjs via node with --platform. From Unibean9/harness-skills.
not rated 18
changed 20d ago A
tokens not measured
SessionStart
08Hook Claude Code
One of 3 in hooks.json
Runs when a session starts on startup, resume, clear and compact, executing session-init.mjs via node. From Unibean9/harness-skills.
not rated 18 21d ago A
tokens not measured
Skill Claude CodeCodex
Provides backend domain guidance for API, data, authorization, reliability, and service-boundary decisions. Workflow skills use it when backend context, validation, or expert review is needed. Follow the repository's existing architecture and conventions before applying generic patterns. Not for running tests, Docker…
not rated 18
changed 24d ago A 71 tokens
Skill Claude CodeCodex
Turns unclear or undecided intent into an agreed direction with enough reasoning and acceptance criteria for planning. Use for fuzzy feature requests, material product, UX, API, architecture, or stack decisions, and bug fixes where more than one cause-aligned remedy remains viable. It does not implement, decompose…
not rated 18
changed 20d ago A 78 tokens
hs-build
11Skill Claude CodeCodex
Implement a settled plan, GitHub issue, or explicit small change while preserving scope, diagnosing feedback, and producing a coherent verified diff. Use when the user asks to implement, build, code, fix an approved task, or pick up issue.
not rated 18
changed 20d ago A 51 tokens
Skill Claude CodeCodex
Independently review a local diff or pull request for concrete reasons not to trust the change. Compare intent and acceptance criteria with current repository behavior, inspect changed tests, apply risk-based domain lenses, and report evidence-backed findings or a clean review. Use after implementation or before…
not rated 18
changed 24d ago A 80 tokens
hs-devops
13Skill Claude CodeCodex
Provides project-aware platform and release-engineering guidance for infrastructure, CI/CD, containers, observability, secrets, and deployment safety. Use when a change touches Terraform or cloud resources, pipeline workflows, Docker or Kubernetes configuration, health signals, identity, or runtime operations. It…
not rated 18
changed 24d ago A 84 tokens
Skill Claude CodeCodex
Design and build production-grade frontend interfaces, context first - capture PRODUCT.md (strategy) and DESIGN.md (visual system), pull reusable tokens into a design system, and apply shared UX/UI rules (color, typography, layout, motion, accessibility, anti-patterns) for brand and product surfaces. Its capabilities…
not rated 18
changed 24d ago A 175 tokens
hs-plan
15Skill Claude CodeCodex
Turns an agreed Decision Brief or clear direction into a grounded implementation-ready plan with independently verifiable tasks and acceptance criteria. Use after hs-brainstorm or when the implementation direction is already settled. It does not choose approaches or write implementation code. With --gh, it publishes…
not rated 18
changed 20d ago A 68 tokens
hs-ship
16Skill Claude CodeCodex
Reconcile an exact reviewed release candidate with live repository policy, then publish or merge only the outward actions the user has authorized. Use when opening or updating a PR, addressing PR review, waiting on required checks, merging, or finishing a landed change. It does not design CI/CD or cloud deployment…
not rated 18
changed 24d ago A 72 tokens
hs-test
17Skill Claude CodeCodex
Provide executable feedback on changed behavior with the smallest relevant white-box or black-box probe, then classify the result and report real evidence. Diagnose failures against accepted intent, domain rules, and current implementation before deciding whether code, tests, fixtures, or environment are wrong. Use…
not rated 18
changed 20d ago A 84 tokens
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: