Skill Claude CodeCodex
Decision guide for delegating to caveman-style subagents. Tells the main thread WHEN to spawn cavecrew-investigator (locate code), cavecrew-builder (1-2 file edit), or cavecrew-reviewer (diff review) instead of doing the work inline or using vanilla Explore. Subagent output is caveman-compressed so the tool-result…
Skill Claude CodeCodex
Ultra-compressed commit message generator. Cuts noise from commit messages while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when "why" isn't obvious. Use when user says "write a commit", "commit message", "generate commit", "/commit", or invokes /caveman-commit.…
Skill Claude CodeCodex
Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman-compress FILEPATH or…
Skill Claude CodeCodex
Quick-reference card for all caveman modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /caveman-help, "caveman help", "what caveman commands", "how do I use caveman".
Skill Claude CodeCodex
Ultra-compressed code review comments. Cuts noise from PR feedback while preserving the actionable signal. Each comment is one line: location, problem, fix. Use when user says "review this PR", "code review", "review the diff", "/review", or invokes /caveman-review. Auto-triggers when reviewing pull requests.
Skill Claude CodeCodex
Show real token usage and estimated savings for the current session. Reads directly from the Claude Code session log — no AI estimation. Triggers on /caveman-stats. Output is injected by the mode-tracker hook; the model itself does not compute the numbers.
Skill Claude CodeCodex
Ultra-compressed communication mode. Cuts output tokens 65% (measured) by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens"…
Plugin Claude Code
Plugin marketplace listing 1 plugin: haytham.
Plugin Claude Code
Idea-to-MVP pipeline: validate a startup idea, scope the MVP, design the architecture, and generate implementation-ready specs.
Instructions file CodexOpenCode
Instructions for arslan70/haytham, covering agents.md, collaboration stance, constitution, guiding principles and meta-system design.
Instructions file
Instructions for arslan70/haytham: See AGENTS.md for all project instructions, collaboration stance, and design decisions.
Agent
Analyze build-vs-buy decisions, produce architecture decisions, and generate pre-implementation research directives for non-standard capabilities. Fires once during Phase 3 (HOW) after the capability model is approved. Context: Phase 2 gate just passed; capabilities.json, system-traits.json, and mvp-scope.md are on…
Agent
Adversarial review of the capability model against the approved MVP scope. Finds missing capabilities and acceptance criteria implied by the scope. Use during Phase 2 (WHAT) after the capability model is produced, before Gate 2.
Agent
Transform MVP scope into a structured capability model and classify system traits. Use during Phase 2 (WHAT) after MVP scope is defined.
Agent
Research competitors for a startup idea using web search. Covers competitor profiles, user sentiment, positioning, switching dynamics, and competitive gaps. Use during Phase 1 (WHY) after idea analysis is complete. Runs in parallel with market-researcher.
Agent
Comparatively score all scout problem candidates on 5 dimensions in a single pass, under a hard web-search budget, and attach a solution angle to each. Fires after idea-scout and before the deterministic select stage of the autonomous scout pipeline. Context: candidates.json exists with 4 candidates; the scout command…
Agent
Gate a raw startup idea, expand it into a structured concept, and extract the concept anchor that anchors all downstream Haytham phases. Fires at the start of Phase 1 (WHY) before any market research runs. Context: Founder pasted a raw idea; no prior session state exists. user: "/haytham:validate a gym community…
Agent
Extract up to 5 problem-anchored candidates from a normalized harvest of pain sources (low-star app reviews, Stack Exchange questions, HN pain comments, Ask HN) across the founder's interest domains. Fires as the first LLM stage of the autonomous scout pipeline, after scoutharvest.py has written harvest files and…
Agent
Research the market landscape for a startup idea using web search. Covers market intelligence (JTBD, sizing, trends, risks). Use during Phase 1 (WHY) after idea analysis is complete. Runs in parallel with competitor-researcher.
Agent
Define MVP scope including core identity, boundaries, success criteria, and user flows. Use during Phase 2 (WHAT) after validation report is approved.
Agent
Synthesize a validation report with a GO, PIVOT, or NO-GO recommendation from completed research findings. Fires once during Phase 1 (WHY) after the founder has reviewed the research brief. Context: research-briefer has finished and the founder approved the brief at Step 4. user: "looks good, continue to the…
Agent
Compile a neutral, fact-based research brief from idea analysis and market research for founder review. Use during Phase 1 (WHY) after market research is complete.
Agent
Generate an implementation-ready OpenSpec directory tree (config.yaml, project.md, specs//spec.md) from upstream Haytham artifacts. Fires once during Phase 4 (SPECS) after architecture decisions are approved. Context: Phase 3 gate just passed; capabilities, system traits, architecture decisions, and build/buy analysis…
Command
Set up a new project from Phase 4 specs, ready for implementation.