Demerzel

Use when a pull request needs review by a different AI model than the one that authored it, or when configuring cross-model code review pipelines in CI/CD.

Run Demerzel's autonomous driver — full cycle or individual phases across all repos

Capture meaningful session state (current cursor, in-flight work, live hypotheses, open questions, do-NOT-carry-forward, success criteria) to state/digests/latest.md so the next session — including one after auto-compaction — can re-enter without re-discovering context cold. Distinct from /learnings (which captures surprises). Validates against docs/contracts/digest-schema.json.

Autonomous research scheduler — selects the most under-researched department by coverage ratio, generates questions from completeness gaps, cross-model validates, auto-produces courses at T>=0.8, rate-limited to 6 cycles/day with kill switch

Personalized AI tutor — turns any topic into a customized, progress-tracked learning course. Use when the user wants to LEARN a concept, skill, language, tool, or domain. Creates a MISSION, curated RESOURCES, iterative lessons with self-checks, and learning-records that persist across sessions. NOT for interrogating the user about their own design (use brainstorming/IDSD for that).

Run one bounded supervised autonomous cycle on the Demerzel governance repo. Picks the smallest unchecked governance slice inside allow_edit, implements it, runs the oracle (`pwsh scripts/verify.ps1`), emits cycle evidence, and stops. Refuses to run unless dev-process-overseer reports loop-eligible and the preflight passes. Use when asked to "run one supervised loop", "advance governance autonomously", or "do one cycle and stop".

Pause every /auto-optimize loop across the GuitarAlchemist ecosystem (GA, ix, Demerzel, tars) by writing the cross-repo HALT-ALL marker per the contract in ga/docs/contracts/2026-05-16-overseer-halt-marker.contract.md. Use halt/resume/status subcommands. Implementation is scripts/demerzel_halt.py.

Self-improvement reflex. When the user corrects an approach ("no, don't do that", "we discussed this before", "stop X"), captures the lesson as a permanent project rule appended to CLAUDE.md so the pattern doesn't repeat in this or future sessions. Cherny called this "the most important loop" in his 2026 Sequoia talk.

Run a meta-compounding cycle — scan evolution log, detect promotions/demotions, assess governance effectiveness, propose self-improvements

Emit nightly per-artifact effectiveness deltas as JSONL. Run as cheap continuous measurement so stagnation surfaces before it accumulates into 27-day staleness. Companion to /demerzel compound — never proposes changes, only measures.

Ask a music theory or guitar question to the Guitar Alchemist AI agent. Routes to the appropriate tool bundle (theory, technique, tab, composer, ga-musician, teacher, coach) and returns a grounded answer. Use when the user asks "ask GA X", "what does the guitar agent say about X", or any music theory/guitar question directed at the GA ecosystem. Supports demo mode: invoke with "demo" or "examples" as the arg to see 3 curated example Q&As covering beginner theory, intermediate improvisation, and technique/injury scenarios.

Control Prime Radiant directly — send commands, add GIS pins/paths, open panels, navigate nodes, and validate state via HTTP API on port 5176.

Teach governance concepts visually through Prime Radiant — map constitutions, beliefs, signals, and protocols to GIS pins, paths, and planet surfaces

Discord-triggered governance team dispatch — parse team requests from Discord, validate, spawn teams

Spawn and manage Demerzel governance agent teams — spawn from template, check status, pause work, dissolve team

Use when exposing localhost dev servers to the internet, sharing demos externally, setting up Cloudflare tunnels, or when user says "set up cloudflare", "share my localhost", "expose demos", "tunnel". Also use when Vite shows "host not allowed" errors.

Phase 3 visual perception loop — generate UI targets, diff against actual renders, and auto-fix iteratively until convergence.

Use when onboarding to an unfamiliar codebase, auditing documentation quality, or building a knowledge map of what is known, unknown, dead, or contradictory in a repo. Works on Java/Spring, TypeScript/React, Rust, .NET, YAML governance repos.

Use when artifact count outpaces productive value, repos feel bloated, unused files accumulate, or ERGOL/LOLLI ratio needs measurement and correction. Works on any repo.

Use when generating documentation from code structure (tree-sitter, LSP, static analysis). Prevents AI slop — docs that restate what code already says. Enforces intention over description.

Visual perception loop — detect errors, screenshot, score, and iteratively fix AI-generated UI components until ship-ready.

Bootstrap recursive self-governance in a consumer repo -- copies kit, customizes constitution, configures Galactic Protocol, sets up algedonic channel, runs initial audit

Reflect on external content (LinkedIn posts, articles, repos) — analyze, cross-reference with Demerzel ecosystem, extract lessons, compound learnings.

Detect and fix drift between documentation and reality — README counts, CLAUDE.md descriptions, cross-repo references, and grammar section counts.

Articulate clearly — transform vague ideas, complex systems, or fuzzy plans into crisp, specific, actionable communication. The anti-BS generator.

Agent health self-check — detect context exhaustion, action loops, wrong repo, git failures, task drift, and confidence miscalibration before they cascade.

Demerzel governance coordinator — dispatcher for all governance subcommands (audit, recon, directive, promote, evolve, teach, loop, patterns, confidence, drive)

Autonomous research scheduler — run one Seldon Plan cycle (question generation, cross-model investigation, novelty detection, course production, compounding)

Factory of factories — given a domain description, bootstrap an entire department with grammar, weights, research config, course pipeline, persona mapping, and behavioral tests. One command → complete production capability.

Decode BS — run text through structural, content, and rhetorical analysis to produce a FuzzyEnum<BsLevel> score with tetravalent mapping

Generate governance dashboard reports with stories, drill-downs, and multi-format output — terminal, Discord, or markdown file

Audit context window consumption across skills, rules, MCP tools, CLAUDE.md, and memory. Surface bloat and recommend optimizations. Inspired by ECC context-budget.

Meta-fix — don't just fix the problem, fix the system that allowed the problem. Generalize fixes into patterns, policies, and prevention.

Run a Streeling University automated research cycle — select department, generate question, investigate, produce course material, compound learnings

End-to-end university course production — research findings to published multilingual course material with review gates

List all Demerzel and Seldon governance commands with descriptions

Run Demerzel governance audit — three levels of validation from schema checks to full governance integrity assessment

Execute Demerzel's three-tier reconnaissance protocol on a target repo — self-check, environment scan, situational analysis

Create governance polls in GitHub Discussions when community input would improve decision quality — used sparingly for genuine decisions

Deep research using ChatGPT + NotebookLM + domain knowledge — cross-model validation for high-quality answers and continuous improvement