prompt-evaluation-runn…
Use when evaluating prompts, LLM outputs, red-team suites, or model behavior with local eval configs and safe provider/cost controls.
Use when apx commands fail or assets are missing from listings, MCP servers fail to start, builds hit command-not-found or module-not-found errors, generated assets look stale, or setup works on one machine but not another.
$ npx -y skills add yeaight7/agent-powerups --skill environment-doctor --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/environment-doctorContext preview
The summary Claude sees to decide when to auto-load this skill.
Use when apx commands fail or assets are missing from listings, MCP servers fail to start, builds hit command-not-found or module-not-found errors, generated assets look stale, or setup works on one machine but not another.
name: environment-doctor description: Use when apx commands fail or assets are missing from listings, MCP servers fail to start, builds hit command-not-found or module-not-found errors, generated assets look stale, or setup works on one machine but not another.
Systematically diagnose why the agent environment, project tooling, or Agent Powerups installation is not working as expected.
Run each and note missing or wrong-version ones:
node --version # 18+ for most Agent Powerups tooling npx --version python --version # 3.9+ for validate scripts git --version apx --version # Agent Powerups CLI docker --version # only if using Docker-based MCP configs
apx list apx validate skills apx validate catalog
If `apx` is not found:
For each MCP config that should be active:
apx mcp check <name> # prerequisites: required commands, env vars apx mcp smoke <name> # launch test: does the server actually start?
Common failures:
| Symptom | Likely cause | Fix | |---|---|---| | `docker: command not found` | Docker not installed | Install Docker | | `docker: not running` | Docker daemon stopped | Start Docker Desktop | | `GITHUB_TOKEN not set` | Missing env var | Set in shell profile | | Server starts then crashes | Package not cached | Run `npx -y <package>` once | | `${REPO_ROOT}` literal in args | Placeholder not replaced | Substitute actual path in config |
python scripts/validate-catalog.py python scripts/validate-skills.py
Each error line names the asset and the exact failing field. Fix in order.
git status --short git diff --stat
Stale generated files will show as modified. Re-run the generator script to bring them current.
List what is expected versus what is set — show presence, not values:
# bash/zsh
printenv | grep -E "GITHUB|CONTEXT7|OPENAI|ANTHROPIC" | sed 's/=.*/=SET/'
# PowerShell
Get-ChildItem Env: | Where-Object { $_.Name -match "GITHUB|CONTEXT7" } | Select-Object Namels <agent-root>/skills/ ls <agent-root>/commands/ ls <agent-root>/mcp/
If paths are empty or missing: verify `APX_REPO_ROOT` or the install path passed to `apx`.
ENVIRONMENT DOCTOR REPORT Platform: <claude-code / codex / gemini-cli / generic> Agent root: <path or "not set"> REQUIRED COMMANDS: node: 20.x ✓ / MISSING ✗ npx: ✓ / MISSING python: 3.11 ✓ / MISSING git: ✓ / MISSING apx: 0.x ✓ / MISSING MCP CONFIGS: <name>: check OK / FAIL — <reason> <name>: smoke OK / FAIL — <reason> ASSET VALIDATION: catalog: PASS / FAIL — [N errors] skills: PASS / FAIL — [N errors] STALE GENERATED ASSETS: none / [list] MISSING ENV VARS: none / [list of names only, no values] BLOCKERS (must fix before proceeding): - [item] NON-BLOCKING: - [item]
Inspired by `Yeachan-Heo/oh-my-codex` `skills/doctor/SKILL.md`. Rewritten in Agent Powerups style with multi-platform coverage for Claude Code, Codex, Gemini CLI, and generic shell environments.
Curated power-ups for coding agents: skills, slash commands, MCP configs, hooks, AGENTS.md templates, and workflows for serious software engineering. Claude Code, Codex, Antigravity CLI, Cursor and more
Repo: yeaight7/agent-powerups
Use when evaluating prompts, LLM outputs, red-team suites, or model behavior with local eval configs and safe provider/cost controls.
Use when creating or reviewing red-team eval plugins, attack templates, grader rubrics, safety fixtures, or model-risk test metadata.
Use when designing, running, debugging, or hardening deterministic eval suites for agent skills, prompts, tool workflows, or MCP-backed cases.
Use when designing tool definitions for a new agent or subagent, an agent shows high retry rates, ambiguous tool invocations, or silent failures, or an…
Use when routing a prompt to a local provider CLI for a second opinion, review, or plan -- you are about to call a provider directly, need the response saved…
Use when starting work in an unfamiliar area of a codebase, spawning a subagent that needs targeted file context, a first search pass missed the relevant file,…