Stats
Actions
Available In
Tags
Plugin
odin
Enforce atomic Git commits with branchless workflow, automated code review, dependency upgrades, and security audits. Debug, refactor, and optimize through structured agentic loops with parallel analysis, verification gates, and auto-revert on regression. Persist conventions, taste preferences, and memory from session transcripts into durable config files.
What's Inside
Skills73
agnix
/agnix
Native agent-configuration lint pass for ODIN and Claude Code surfaces: skill frontmatter, instruction memory, hooks, MCP servers, plugin manifests, and subagent files. Use when lint agent config, validate SKILL.md, check CLAUDE.md, check hooks, check MCP config, agnix.
ai-collab-protocols
/ai-collab-protocols
Surface in-task-collaboration protocols when the user describes an AI workflow informally — URL-as-entity-reference, PR-comment threads as session memory. Trigger when the user names entities by colloquial label instead of stable URL, asks "how should I structure this for Claude", or describes a multi-step Claude workflow without a durable handle. Apply reactively, not as a checklist.
askme
/askme
Verbalized Sampling (VS) protocol for intent exploration before planning, mode-aware. Default `exhaustive` runs full VS; `collaborative` runs tip-sharing dialogue; `adversarial` walks the design tree one fork at a time. Auto-detects from phrasing ("help me refine" → collaborative, "poke holes" → adversarial); override via `/askme adversarial|collaborative|exhaustive`. Use for ambiguous tasks or maximum clarifying questions before committing.
banthis
/banthis
Banthis extends project memory with hard user prohibitions by writing a managed AGENTS.md/CLAUDE.md section and preserving all surrounding content. Use when the user says "stop doing X", "never do that again", "ban this", "remember not to", or "add a project rule".
repo-intel
/repo-intel
Native repository intelligence from git history, codegraph when indexed, manifest reads, and compressed repo digests; no cache, no generated state, every signal is recomputed on demand. Use when the user asks to analyze git history, find hotspots, code coupling, bus factor, bugspots, code ownership, entry points, or repo health.
MCP Servers4
Output Styles6
Linus
linus
Linus Torvalds. Special-case elimination is good taste. Consumer contracts are covenant. Complexity discipline is law. Blunt about the work; be harsh.
Eval
benchmark
Margin Eval runner style. Disables token-conservation heuristics, enforces complete and un-truncated code blocks for AST parsers, and suppresses conversational filler.
Builder
builder
Communication style for non-technical builders — product managers, founders, designers, and no-code/low-code users who build things without deep programming expertise. Leads with outcomes, uses plain language, maintains honesty without jargon overload.
ODIN
odin
Outline Driven Development(ODIN)'s approach to Claude output style; Precise and effective
Duet
duet
Output style for the duet working posture (user as director, agent as executor). Minimizes cognitive load between picks: decisions before prose, structural/taste framing first, jargon on demand, batched questions with concrete previews, short execution updates, no validation language, no recap. Goal — eliminate the review bottleneck and prevent codebase-understanding debt by distributing review across the task at pick-time.
Stats
Version1.15.52
LanguagePython
Stars27
Copy clicks5
MaintenanceExcellent
LicenseApache-2.0
Last Commit
Added
Available In
