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 skills/benjaminthomas/spec-driven-dev/spec-verifynpx skills add benjaminthomas/spec-driven-dev --skill spec-verifygit clone --depth 1 https://github.com/benjaminthomas/spec-driven-devWhat 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 | $0.00210 | $0.01927 |
| Opus 5 | $0.00105 | $0.00963 |
| Sonnet 5 | $0.00042 | $0.00385 |
| Haiku 4.5 | $0.00021 | $0.00193 |
Grade C, and why
spec-verify scanned grade C with 2 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.
Tells the agent never to refusehighAnti-refusal
Suppressing the ability to decline removes a core safety control; a later harmful request then succeeds.
If the user asks to verify a feature that wasn't built through `spec-create`, don't refuse — Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
Check whether the dev server is already running (e.g. `curl -sf http://localhost:3000` or the How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verify Feature
Close the loop that spec-create → spec-implement leaves open: passing lint, typecheck, and
build proves the code compiles, not that the feature does what the spec says it does. This skill
reads a spec's acceptance criteria and checks each one against the actual running application —
clicking through the UI, calling the real API routes, or inspecting real database state produced
by the real code path. It produces a report the user can trust before shipping.
When there's no spec folder
If the user asks to verify a feature that wasn't built through spec-create, don't refuse —
interview them briefly for what was built and what "working" means (the golden path, the edge
cases they care about), then skip to Step 3 treating their
answers as the criteria list. The spec-driven path below is the common case, not the only one.
Step 1: Locate the spec
- If the user named a feature, use
specs/{feature}/. - Otherwise look for a
specs/*/README.mdwhose Task Status section is fully checked (all tasks- [x]) but has noverification-report.mdsibling yet — that's the most recently completed, not-yet-verified feature. - If
spec-implementhasn't finished all batches yet (some tasks still- [ ]), tell the user and ask whether to verify what's done so far or wait for the rest — verifying half a feature can produce confusing false failures for criteria that depend on an unbuilt task.
Step 2: Build the master criteria checklist
Read requirements.md's Acceptance Criteria section (feature-level, the whole thing working
end-to-end) and every tasks/task-*.md's own Acceptance Criteria section (task-level). Merge
them into one ordered list, each item tagged with its source (requirements or the task
filename).
Split the merged list into two kinds, because they're verified completely differently:
- Static criteria — things like "
npm run checkpasses" or "npx next buildsucceeds."spec-implement's review gate already ran these during the build. Re-run them once quickly to confirm nothing regressed since (uncommitted fix-ups, a rebase, etc.), but don't treat this as the interesting part of verification. - Behavioral criteria — anything that describes what a user, an API caller, or a scheduled job actually observes ("a FREE user sees a Renew button," "the cron doesn't double-send," "removing a member revokes their access on next request"). These are why this skill exists — they require driving the real app, not reading the diff.
What ships with it
1 file 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.
- yesterday First seen · 141 lines · 210 tokens per session scan C 326efa28505f
spec-verify is a skill published in the GitHub repository benjaminthomas/spec-driven-dev (1 stars, last pushed 28d ago), licensed MIT. It adds 210 tokens to every session and 1,927 once invoked, about $0.0011 per session on Opus 5. A static security scan graded it C with 2 findings (tells the agent never to refuse, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
academic-paper
12-agent academic paper writing pipeline. 11 modes (full/plan/outline/revision/revision-coach/abstract/lit-review/format-convert/citation-check/disclosure/rebuttal-audit). 6 paper types, 5 citation formats, bilingual abstracts, LaTeX/DOCX-via-Pandoc/PDF output. Style Calibration + Writing Quality Check + Anti-Patterns…
academic-paper-reviewer
Multi-perspective academic paper review with dynamic reviewer personas. Simulates 5 independent reviewers (EIC + 3 peer reviewers + Devil's Advocate) with field-specific expertise. Supports full review, re-review (verification), quick assessment, methodology focus, Socratic guided, and calibration modes. Triggers on…
agent-platform-alert-configuration
Configures best-practice alerting policies for Google Cloud Vertex AI / Agent Platform agents on Agent Runtime. Use when analyzing, writing, or deploying alerting policies to monitor agent latency, error rates, and quality metrics (response quality, tool use, hallucination). Also use when provisioning online monitors…
agent-platform-eval-flywheel
Measures and improves the quality of AI models and agents on Google Cloud using the Eval Quality Flywheel methodology. Use when evaluating an agent or model, building an eval dataset, picking or writing evaluation metrics, analyzing failures, comparing results before and after a fix, or when guidance is needed on…
agent-platform-inference
Connects to and performs inference with Google Cloud Agent Platform GenAI models, including First-Party Gemini models and Third-Party OpenMaaS models (Llama, DeepSeek, Qwen, etc.). Use when you need to generate code for calling Gemini or OpenMaaS models, authenticate with GenAI SDK, OpenAI SDK, or legacy Agent…
gemini-omni-flash-api
Use this skill for generative video editing, text-to-video, image-referenced video generation, and first-frame-to-video transition animations using the official google-genai SDK. Includes workflows for pre-processing/optimizing high-resolution or long source videos with ffmpeg, stripping audio for full sound…