Reduce, observe, reason, predict, probe, then fix — no guessing.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Systematic Debugging Protocol
You are now operating under Dave's debugging protocol. This is not optional guidance — it is the mandatory sequence for investigating any failure. Agents who skip steps reliably produce wrong fixes that waste time and create new bugs.
## The Protocol
Every debug cycle follows this sequence. No skipping steps.
### 1. REDUCE
Before tracing anything, reduce to the **smallest case that still fails**.
- Remove variables one at a time until you find the one that matters
- If a complex test fails, try the simplest version first
- If multiple things are broken, isolate one failure before investigating
- The dimension you add back that triggers the failure is where the bug lives
If the minimal case passes, use **divergence-point analysis**: compare the passing case against the failing case signal-by-signal, log-by-log, output-by-output. The first thing that diverges is closest to root cause.
### 2. OBSERVE
What does the data show? Not what you think should happen — what **actually** happens.
- List specific values, states, outputs at specific points
- Cover ALL relevant signals/variables, not just the one that looks wrong
…
The line at the top of this page pins the label git:20260709.887cd3f, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
GET https://markdownregistry.com/api/v1/artifacts/art_edc7wwvhqgvchg2v
GET https://markdownregistry.com/api/v1/resolve?ref=cercano-ai/cercano/systematic-debugging
GET https://markdownregistry.com/api/v1/blob/307a58b3207f8ecefbdde12530ee3695869aac5d98da3e2a5c9ce2a77e62ccdc
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
cercano-ai/cercano · .agents/skills/cercano-classify/SKILL.md · Classify or triage text using the configured model route via Cercano. Returns a category, confidence level, and brief…
cercano-ai/cercano · .agents/skills/cercano-config/SKILL.md · Query or update Cercano's runtime configuration without restarting the server. Use action 'get' to list available local…
cercano-ai/cercano · .agents/skills/cercano-deep-research/SKILL.md · Deep multi-source research tool that identifies authoritative sources, systematically searches, analyzes and ranks…
cercano-ai/cercano · .agents/skills/cercano-document/SKILL.md · Generate doc comments for exported Go symbols using routed AI and write them directly to the file. Cercano handles the…
cercano-ai/cercano · .agents/skills/cercano-explain/SKILL.md · Explain code or text using the configured model route via Cercano. Returns a clear explanation of what the code does…
cercano-ai/cercano · .agents/skills/cercano-extract/SKILL.md · Extract specific information from text using the configured model route via Cercano. Returns only the relevant sections…
cercano-ai/cercano · .agents/skills/cercano-fetch/SKILL.md · Fetch a URL and extract readable text content locally. Returns full extracted text (not a summary) — HTML is stripped…
cercano-ai/cercano · .agents/skills/cercano-init/SKILL.md · Initialize Cercano for a project by scanning the repo and building a context file. Makes all Cercano tools…
cercano-ai/cercano · .agents/skills/cercano-local/SKILL.md · Run prompts through Cercano with configured routing. Handles chat-style queries and code generation with validation…
cercano-ai/cercano · .agents/skills/cercano-models/SKILL.md · List AI models available on the active local runtime configured in Cercano. Returns model names, sizes, and…
cercano-ai/cercano · .agents/skills/cercano-research/SKILL.md · Research a question using DuckDuckGo search and routed AI analysis. Crafts search queries, fetches top results, and…
mvschwarz/openrig · skills/_canonical/process/systematic-debugging/SKILL.md · Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
mvschwarz/openrig · packages/daemon/specs/agents/shared/skills/process/systematic-debugging/SKILL.md · Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
cbrock84/headcount · plugins/technology/skills/systematic-debugging/SKILL.md · Finds the root cause of a bug, test failure, or unexpected behavior before proposing any fix. Use this whenever…
a5c-ai/babysitter · library/methodologies/superpowers/skills/systematic-debugging/SKILL.md · Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes. Requires root cause…