AGENT SKILLS · MISSION KIT
Useful agent skills ranked by real adoption
Browse verified skills from public repositories. Compare what each skill does, package contents, install activity, and source evidence before adding it to your agent.
- CATALOG SCOPE
- VERIFIED
- skills
- 22,811
- repositories
- 3515
- last synced
- Nov 2, 2026
jcogermobile-scaffold
- The day-one gate for a new React Native / Expo app. Make the decisions that are cheap at scaffold time and brutal at day 45. Use when creating a new RN/Expo project, when bootstrapping a repo before the first feature, or to retrofit-check an existing project for missing foundations (toolchain pins, component-library commitment, sheet/keyboard architecture, quality tooling, EAS env binding, store-compliance scaffold, analytics and crash reporting, notification and entitlement plumbing). The ope
- 1h installs
- 1
- GitHub stars
- 1
- Updated
- Oct 1
jcogermotion-direction
Art direction for website motion. Use when deciding how a site or page should move, choosing a motion personality, writing a motion spec before any animation is built, or auditing the motion on a live page. Trigger phrases include "motion direction", "how should this page move", "animation direction", "motion personality", "motion spec", "the animations feel random", "audit the motion on this site".
- 1h installs
- 1
- GitHub stars
- 1
- Updated
- Oct 1
jcogerweb-build
The build bridge. Converts approved direction specs (LAYOUT SPEC, PAGE SPEC, MOTION SPEC, CONVERSION SPEC, site maps, BEAT SHEETs) into a BUILD SPEC and a kickoff prompt that produce an authored first build instead of a generic one. Use when the user says "spec the build", "get this ready for Claude Code", "write the tokens", "write the kickoff prompt", "stack decision", or "check the build against the spec".
- 1h installs
- 1
- GitHub stars
- 1
- Updated
- Oct 1
johnkimeccompound
After a task, write one standing project lesson into AGENTS.md so the next agent starts with it. Use for /compound or when a finished task revealed a command, path, or constraint the next task will need. Skip notes that only describe where this task stopped.
- 1h installs
- 1
- GitHub stars
- 1
- Updated
- Oct 9
johnkimecproject-doc
Write a project doc with the user before any design is settled or any code is written: problem and scope, breakdown, decisions the user makes, then an implementation plan. Use for /project-doc, when starting a new project or design, when scoping a project before building, or when returning to an in-progress project doc in docs/.
- 1h installs
- 1
- GitHub stars
- 1
- Updated
- Oct 9
johnkimecwalkthrough
Explain a codebase in one linear document whose code snippets were printed by a command. Use Showboat note for commentary and Showboat exec with sed, grep, or cat for snippets. Use for /walkthrough, \"walk me through this code\", or \"how does this actually work\".
- 1h installs
- 1
- GitHub stars
- 1
- Updated
- Oct 9
crottoloflowise-api
Manages Flowise AI instances via REST API. Performs CRUD operations on chatflows, agentflows (V2/V3), assistants, custom tools, variables, and document stores. Sends predictions (chat messages) with streaming, file uploads, and human-in-the-loop support. Queries vector stores and manages feedback/leads. Uses bearer token authentication. Triggers on "flowise", "chatflow", "agentflow", "prediction", "send message to flow", "document store", "vector upsert", "flowise API".
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Mar 14
harisjavaid85cross-check-with-codex
Cross-checks an identified artifact with Codex through an independent review loop, reconciles material findings, and preserves a local audit record. Use when the user asks Claude to check, cross-check, review, or get a second opinion from Codex on a plan, diagnosis, design, implementation, or other artifact.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85diagnosing-bugs
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85domain-modeling
Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a CONTEXT.md, or recording or editing an ADR.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85grill-with-context
A relentless interview to sharpen a plan or design against the project's domain model, which also creates docs (ADRs and glossary) as we go.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85handoff
Compact the current conversation into a focused handoff document for another agent. Use when the user wants to continue the work in a fresh session or transfer it to another agent.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85kb-ingest
Ingest sources into a named file-based knowledge base: normalize each to Markdown and register it under a stable key. Use when the user asks to add papers, PDFs, URLs, or notes to a specific KB, or to ingest sources into a KB for review before integrating them.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85research
Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85resolving-merge-conflicts
Use when you need to resolve an in-progress git merge/rebase conflict.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85scaffold-exercises
Create exercise directory structures with sections, problems, solutions, and explainers that pass linting. Use when user wants to scaffold exercises, create exercise stubs, or set up a new course section.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85to-done
Drain a spec's agent-ready tickets to a reviewed draft PR, running each ticket unattended in a sandboxed container.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85to-questionnaire
Turn a decision you can't fully answer into a questionnaire for someone else to fill in.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85wait-what
Stop. That last message did not land: re-pitch it.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85writing-beats
Writing, exploit; assemble raw material into a journey of beats, grounding each term before a beat leans on it.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
harisjavaid85writing-shape
Writing, exploit: shape raw material into an article, paragraph by paragraph.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 9
lucaswilliameufrasiocross-repo-standard-sync
- Find the canonical owner of a development standard, audit drift, or propagate an approved standard across repositories and agent configurations. Use when the user says their preferred workflow is out of sync, wants common rules synchronized, or asks which copy is authoritative. Separate personal defaults, organization policy, and repository facts; preserve local commands and exceptions, and never overwrite ambiguous differences without approval.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 8
lucaswilliameufrasioexplicit-code-style
- Apply explicit, safe coding conventions when writing or reviewing code: use multiline braced control flow, explicit loops, and strict TypeScript value handling. Use when editing, reviewing, or refactoring application code; respect API contracts and language-specific constraints.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 8
lucaswilliameufrasioperformance-optimization
- Diagnose and improve application performance or resource efficiency using measured evidence. Always use when the user says an app/API is slow, reports a performance regression or bottleneck, asks to optimize/tune code, or wants higher throughput or lower latency, CPU, memory, or cost—even if they request a specific code change before providing a profile. Reproduce the workload, locate the bottleneck, change one cause at a time, and verify comparable runs.
- 1h installs
- 1
- GitHub stars
- 0
- Updated
- Oct 8