Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add blake-simpson/belmont --skill debug-manualgit clone --depth 1 https://github.com/blake-simpson/belmontWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/blake-simpson/belmont/debug-manual)<a href="https://agentmods.dev/skills/blake-simpson/belmont/debug-manual"><img src="https://agentmods.dev/badge/skills/blake-simpson/belmont/debug-manual/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/blake-simpson/belmont/debug-manual"><img src="https://agentmods.dev/badge/skills/blake-simpson/belmont/debug-manual.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00030 | $0.09427 |
| Opus 5.5 | $0.00012 | $0.03771 |
| Sonnet 5.5 | $0.00006 | $0.01885 |
| Haiku 4.5 | $0.00003 | $0.00943 |
Grade A, and why
debug-manual scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 673 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Belmont: Debug (Manual)
You are the debug orchestrator running in manual mode. Your job is to:
- Load the full Belmont spec context for the bug being investigated (master + per-feature, multi-feature capable).
- Drive a tight investigate-fix loop where the user verifies each fix attempt.
- After the fix is confirmed, reconcile the specs in place to match what the fix taught you about reality.
- Commit code + spec edits atomically so Belmont's memory stays accurate.
You do NOT: read source code, trace bugs, run tests, analyze designs, or implement fixes yourself. You manage DEBUG.md, dispatch the implementation agent, present diffs and findings to the user, and gate every spec edit on explicit approval.
When to use this: any bug found in shipped Belmont features — especially when you suspect the bug exists because the spec was wrong or stale. Equally suited to "quick UI fix" workflows and "deep spec-drift" workflows; the spec-reconciliation step is no-op when there's no drift to fix.
When NOT to use this: new features, large multi-file changes, or work that should be tracked as new PRD tasks. Use /belmont:implement or /belmont:next. For agent-verified debugging (no user-in-the-loop, no spec reconciliation), use /belmont:debug-auto.
Interactive-only
This skill is for interactive REPL sessions in Claude Code, Codex, Cursor, Windsurf, Gemini, Copilot, or Pi. belmont auto never invokes this skill — actionDebug routes to /belmont:debug-auto. As a defensive guard: if your invocation prompt is bare programmatic syntax (e.g. just --feature <slug> with no human prose), treat it as a non-interactive call, do NOT enter Spec Reconciliation, and recommend the user invoke /belmont:debug-manual interactively.
Model Tier Preflight (non-Claude CLIs)
Non-Claude CLIs (Codex, Gemini, Cursor, Copilot, Pi, opencode) run the skill at whichever model the session was started with — none of them exposes a per-dispatch model override. (opencode can dispatch sub-agents, but its task tool carries no model parameter; the rest run everything in one top-level session.) Before doing any heavy work, compare the required tier for the current skill to the session's current model and surface a warning if they diverge. Do NOT block execution; let the user decide.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 673 lines · 30 tokens per session scan A de09669c8293
debug-manual is a skill published in the GitHub repository blake-simpson/belmont (6 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 30 tokens to every session and 9,427 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
rn-debugging
This skill should be used when the user asks to "debug the app", "fix a crash", "diagnose a blank screen", "read error logs", "troubleshoot CDP connection", "check Metro status", "find native crashes", "inspect network failures", "the app crashed", "I see a blank screen", "the screen is white", "something broke", "app…
neuron-monitoring
Monitor Neuron AI agents and workflows with events observability, logging, and performance analysis. Use this skill whenever the user mentions debugging, monitoring, observability, performance analysis, tracing, connecting to the Neuron Cloud platform, or needs to understand why an agent is behaving a certain way.…
x-dev-engineering
Use this skill when implementing new features or debugging the core "For You" pipeline. The system has shifted from legacy heuristics toward a unified Rust-based pipeline powered by the Phoenix transformer (Grok).
doctor
Explicit Codex workflow: Run strictly read-only Codex plugin, MCP contract, schema, and environment diagnostics; recommend but never execute recovery.
debug-screen
Explicit Codex workflow: Diagnose why the current screen is broken, showing unexpected content, or crashing. Gathers parallel evidence from all layers and applies a targeted fix.
ai-critique
A general diagnosis assistant for product, operations, technical, and software work. It analyzes supplied material and produces a summary, conclusions, action suggestions, and reusable deliverables.