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 agentmods add commands/dinhnguyenngoc/spec-driven-claude-code/inspectgit clone --depth 1 https://github.com/dinhnguyenngoc/spec-driven-claude-codeWrote 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/commands/dinhnguyenngoc/spec-driven-claude-code/inspect)<a href="https://agentmods.dev/commands/dinhnguyenngoc/spec-driven-claude-code/inspect"><img src="https://agentmods.dev/badge/commands/dinhnguyenngoc/spec-driven-claude-code/inspect.svg" alt="Measured on agentmods" 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 | $0.00023 | $0.02438 |
| Opus 5 | $0.00012 | $0.01219 |
| Sonnet 5 | $0.00005 | $0.00488 |
| Haiku 4.5 | $0.00002 | $0.00244 |
Grade A, and why
inspect 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 4d 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 — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/inspect — Interrogate Current State
"Answer with evidence, not memory."
Purpose
Answer a specific question about the software's current state ("does feature X exist/work?", "how is Y configured?", "does a newly added bookmark get checked for duplicates?") using 3 independent evidence tiers, and detect mismatches among them. Read-only — changes nothing, no gate, no artifact pipeline.
Workspace Mode: if the session root declares
Mode: workspace→ resolve the target repo perCLAUDE.md§Workspace Mode before anything else; every path, probe, and gate below is relative to the target repo, and the workspace disk-check applies at the gate.
Free-text goes down this same path: the user asks a current-state question in plain language (without typing
/inspect) → the orchestrator answers per this command's exact output contract (seeCLAUDE.md§Natural-Language Task Routing)./inspectis only the explicit entry point; the answering discipline is one and the same.
Usage
| Invocation | Result |
|---|---|
/inspect <question> |
Default: Tier 1 (records) + Tier 2 (code) — no running stack required |
/inspect --live <question> |
Adds Tier 3: probe the running artifact (opt-in — requires the stack up, may write throwaway test data) |
/inspect --report <question> |
As above + save the answer to reports/inspect/INSPECT-<slug>.md (opt-in — by default answers only on chat) |
Examples: /inspect does a newly added bookmark get checked for duplicates · /inspect --live is the rate limit on /auth/login working · /inspect is the refresh token rate-limited
Scope Clarification (how it differs from neighboring commands)
| Command | Answers the question | How /inspect DIFFERS |
|---|---|---|
/discover |
"What is this repo?" (whole-repo survey, run once at onboarding) | 1 targeted question, targeted read, run as many times as you like |
/verify |
"Do ALL acceptance criteria pass on the artifact?" (gate, heavy) | No gate, no promote verdict, scope narrowed to exactly the question |
/debug |
"Why is it BROKEN?" (there's an error symptom, find root cause) | No symptom needed — asks about current state, even when everything works fine |
| free-text | Like /inspect but with an ad-hoc format |
Mandatory output contract (3-tier table + evidence rating + mismatch flag + citations) — free-text is routed to the same contract |
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.
- 4d ago First seen · 113 lines · 23 tokens per session scan A eaf72cbcd8ef
inspect is a command published in the GitHub repository dinhnguyenngoc/spec-driven-claude-code (20 stars, last pushed 5d ago), licensed MIT. It adds 23 tokens to every session and 2,438 once invoked, about $0.0001 per session on Opus 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 2026-08-30.
Other commands, from other repositories
brooks-audit
Run a Brooks-Lint architecture audit.
kiro-validate-design
Command "kiro-validate-design" from gotalab/cc-sdd, covering technical design validation, core task, execution steps, important constraints and tool guidance.
sdd-setup
Onboarding wizard — DocLanguage, Memory Bank, architecture snapshot, working agreements (quality gate, TDD working mode).
verify
Perform a non-destructive post-implementation verification gate validating the implementation against spec.md, plan.md, tasks.md, and constitution.md.
fix
⚡⚡ Analyze and fix small issues [AUTO DETECT COMPLEXITY].
extend
Capture a mid-PR sub-idea and implement it onto the current open PR's branch — no new branch, no new PR. Preserves Verify → Review → Deliver.