agent-sh

Use when user asks to 'lint agent configs', 'validate skills', 'check CLAUDE.md', 'validate hooks', 'lint MCP'. Validates agent configuration files against 421 rules across 10+ AI tools.

Use when user asks to 'lint agent configs', 'validate skills', 'check CLAUDE.md', 'validate hooks', 'lint MCP'. Validates agent configuration files against 421 rules.

Use when analyzing code quality

Use when testing valid context and agent configuration

Use when testing custom agent validation

Use when testing per-client skill rules

Use when testing per-client skill rules

Use when testing per-client skill rules

Use when user asks to "discover tasks", "find next task", "prioritize issues", "what should I work on", or "list open issues". Discovers and ranks tasks from GitHub, GitLab, local files, and custom sources.

Use when preparing releases, validating cross-platform compatibility, or updating installation infrastructure. Meta-skill for maintaining AgentSys's 3-platform architecture.

Use when user asks to "run repo intel", "generate repo map", "analyze repo", "query hotspots", "check ownership", or "bus factor". Unified static analysis - git history, AST symbols, project metadata.

Use when mapping code paths, entrypoints, and likely hot files before profiling.

Use when generating performance hypotheses backed by git history and code evidence.

Structured AI debate templates and synthesis. Use when orchestrating multi-round debates between AI tools, 'debate topic', 'argue about', 'stress test idea', 'devil advocate'.

Use when user wants to clean AI slop from code. Use for cleanup, remove debug statements, find ghost code, repo hygiene.

Use when the user asks about plan drift, reality check, comparing docs to code, project state analysis, roadmap alignment, implementation gaps, or needs guidance on identifying discrepancies between documented plans and actual implementation state.

Use when coordinating multiple enhancers for /enhance command. Runs analyzers in parallel and produces unified report.

Use when improving agent prompts, frontmatter, and tool restrictions.

Use when improving CLAUDE.md or AGENTS.md project memory files.

Use when checking cross-file consistency: tools vs frontmatter, agent references, duplicate rules, contradictions.

Use when improving documentation structure, accuracy, and RAG readiness.

Use when reviewing hooks for safety, timeouts, and correct frontmatter.

Use when analyzing plugin structures, MCP tools, and plugin security patterns.

Use when improving general prompts for structure, examples, and constraints.

Use when generating performance hypotheses backed by git history and code evidence.

Use when synthesizing perf findings into evidence-backed recommendations and decisions.

Use when managing perf baselines, consolidating results, or comparing versions. Ensures one baseline JSON per version.

Use when running performance benchmarks, establishing baselines, or validating regressions with sequential runs. Enforces 60s minimum runs (30s only for binary search) and no parallel benchmarks.

Use when mapping code paths, entrypoints, and likely hot files before profiling.

Use when appending structured perf investigation notes and evidence.

Use when profiling CPU/memory hot paths, generating flame graphs, or capturing JFR/perf evidence.

Use when running controlled perf experiments to validate hypotheses.

Use when user asks to "deep review the code", "thorough code review", "multi-pass review", or when orchestrating the review loop in a delivery pipeline. Provides review pass definitions (code quality, security, performance, test coverage), signal detection patterns, and iteration algorithms.

Use when user asks to "prepare delivery", "run quality gates", "review code", "run pre-gates", "deslop and review", "validate before shipping", or has finished implementation and wants quality checks before shipping.

Use when validating test coverage quality for new code. Checks that tests exist, are meaningful, and actually exercise new code - not just path matching.

Use when user asks to "validate delivery", "check readiness", or "verify completion". Runs tests, build, and requirement checks with pass/fail instructions.

Browse and interact with web pages headlessly. Use when agent needs to navigate websites, click elements, fill forms, read content, or take screenshots.

Authenticate to websites with human-in-the-loop browser handoff. Use when user needs to log into a website, complete 2FA, or solve CAPTCHAs for agent access.

Compact conversation transcripts into themed knowledge files.

Analyze accumulated skillers knowledge and suggest skills, hooks, and agents. Use when compacting is done and patterns need classification. Checks existing ecosystem before recommending.

Use when Hermes needs Linux desktop observation or control through the computer-use-linux MCP server.

Use when the user asks about plan drift, reality check, comparing docs to code, project state analysis, roadmap alignment, implementation gaps, or needs guidance on identifying discrepancies between documented plans and actual implementation state.

Creates message queues, workers, job workflows, and fan-out broadcasts using glide-mq on Valkey/Redis Streams. Provides API reference, code patterns, and configuration for queues, workers, delayed/priority jobs, schedulers, batch processing, DAG workflows, request-reply, serverless producers, and AI-native primitives (usage tracking, token streaming, suspend/resume, budget caps, fallback chains, dual-axis rate limiting, rolling usage summaries, vector search, HTTP proxy/SSE). Triggers on "glide-mq", "glidemq", "job queue valkey", "background tasks valkey", "message queue redis streams", "glide-mq LLM queue", "glide-mq AI orchestration queue", "glide-mq token rate limiting", "glide-mq model fallback", "glide-mq human-in-the-loop queue", "glide-mq vector search", "glide-mq AI pipeline".

Use when user asks to "discover tasks", "find next task", "prioritize issues", "what should I work on", or "list open issues". Discovers and ranks tasks from GitHub, GitLab, local files, and custom sources.

Structured AI debate templates and synthesis. Use when orchestrating multi-round debates between AI tools, 'debate topic', 'argue about', 'stress test idea', 'devil advocate'.