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 shawnclybor/clybor-claude-tooling --skill why-diagnosticgit clone --depth 1 https://github.com/shawnclybor/clybor-claude-toolingWrote 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/shawnclybor/clybor-claude-tooling/why-diagnostic)<a href="https://agentmods.dev/skills/shawnclybor/clybor-claude-tooling/why-diagnostic"><img src="https://agentmods.dev/badge/skills/shawnclybor/clybor-claude-tooling/why-diagnostic/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/shawnclybor/clybor-claude-tooling/why-diagnostic"><img src="https://agentmods.dev/badge/skills/shawnclybor/clybor-claude-tooling/why-diagnostic.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 | $0.00100 | $0.03001 |
| Opus 5.5 | $0.00040 | $0.01200 |
| Sonnet 5.5 | $0.00020 | $0.00600 |
| Haiku 4.5 | $0.00010 | $0.00300 |
Grade A, and why
why-diagnostic 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 yesterday.
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 — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Why Diagnostic
Purpose
When something happens in agent or tool work and the user asks "why," the temptation is to produce a tidy causal story: X happened because Y. That story is usually wrong, or right for the wrong reason, because it skips the frame the question is being asked inside.
This skill runs a three-step diagnostic borrowed from Wittgenstein's On Certainty:
- Name the language-game. What practice is the "why" being asked inside? Debugging? Prompt design? Architecture justification? The same event has different "real" causes in each.
- Surface the hinges. What's being held fast — unexamined — for the explanation to even make sense? These are the assumptions doing silent work.
- Name the stop point. Where does the chain of "because" actually give out? Don't fake further depth. State it.
That analytical structure stays intact — those labels (language-game, hinges, chain of because, stop point, the hinge-worth-testing) are the methodology, and they're how the model reasons through the question internally. But the user-facing output is a translation: connected prose in incident-report shape, not labeled blocks. The user never sees the labels; they see a short post-mortem-style write-up that reflects what the labeled analysis produced.
When to use
Trigger when the user asks "why" about anything in the agent/tool workflow:
- A tool failed or behaved oddly — "why did the MCP call return nothing", "why did Brave Search return zero results", "why did this script error"
- An agent's output isn't what expected — "why did the agent format this as a table when I said plain English", "why did it ignore the rubric"
- A design choice needs justifying — "why am I using Langfuse here vs. GitHub Actions", "why should the scorer be flat instead of nested"
Do NOT use for:
- Factual lookups ("why is the sky blue") — just answer
- Emotional/personal "why" questions — that's not what this is for
- Cases where the user just wants the direct fix and the "why" is rhetorical
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.
- yesterday Changed · +30 lines 3b006b4440e2
- 18d ago Changed · +2 lines b9f25cba9bbd
- 2mo ago First seen · 149 lines · 100 tokens per session scan A b4a447a65a60
why-diagnostic is a skill published in the GitHub repository shawnclybor/clybor-claude-tooling (6 stars, last pushed yesterday), licensed MIT. It adds 100 tokens to every session and 3,001 once invoked, about $0.0004 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.
why-diagnostic compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| why-diagnosticshawnclybor/clybor-claude-tooling | Skill | 6 | yesterday | MIT | A | 100 | Claude Code |
| debugSienkLogic/plan-build-run | Skill | 18 | 7mo ago | MIT | A | 13 | Claude Code |
| scanSienkLogic/plan-build-run | Skill | 18 | 7mo ago | MIT | A | 18 | Claude Code |
| audit-fixSienkLogic/plan-build-run | Skill | 18 | 7mo ago | MIT | A | 22 | Claude Code |
| healthSienkLogic/plan-build-run | Skill | 18 | 7mo ago | MIT | A | 13 | Claude Code |
| forensicsSienkLogic/plan-build-run | Skill | 18 | 7mo ago | MIT | A | 23 | Claude Code |
| yd-finguardLeeYudok/agents-scaffold | Skill | 38 | yesterday | MIT | A | 63 | Claude Code |