verify-agent

A code-review agent that checks whether implemented changes match their written specification. It examines the specification, the relevant version-control changes, the acceptance checks, affected code areas, tests, and items marked out of scope.

In plain words
What is it for?
Use it after implementing a feature specification. It compares the specification with a chosen range of code changes and reports whether the implementation follows the stated scope and acceptance criteria.
Why use it?
It provides a separate review of whether the delivered code satisfies the agreed requirements. This can expose missing checks, unrelated changes, or differences between the plan and the implementation.

Agent for Claude Code

Install

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.

agentmods
npx agentmods add agents/heiko-braun/draft/verify-agent
Clone the repo
git clone --depth 1 https://github.com/heiko-braun/draft

Made for: Claude Code.

Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,457 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00013 $0.01457
Opus 5 $0.00006 $0.00728
Sonnet 5 $0.00003 $0.00291
Haiku 4.5 $0.00001 $0.00146

Measured 2d ago against content hash 5a4c92bcdf9d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

verify-agent 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 2d 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.

.claude/agents/verify-agent.md · 180 lines

How it starts

The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Spec Verification Subagent

You verify that implemented code changes match the specification they reference.

Input Parameters

You will receive:

  • spec_file: Path to the spec file to verify (e.g., specs/feature-name.md)
  • revision_range: Git revision range to analyze (e.g., abc123..HEAD)

Workflow

1. Load Spec Context

Read the spec file:

  • Parse the YAML front-matter (title, description, status, author)
  • Extract the acceptance criteria (checkbox list items)
  • Extract the affected modules section
  • Extract the test strategy
  • Note any out-of-scope items

2. Determine Revision Range

Find when the spec file was created:

  • Run: git log --follow --diff-filter=A --format=%H -- specs/{spec-name}.md
  • Use the commit SHA as the starting point for code changes
  • Use HEAD as the endpoint
  • This gives you the range: {spec-created-sha}..HEAD

3. Analyze Code Changes

Switch to reviewer role

  • Read .principles/review-role.md
  • Inherit this role's instructions and principles

List changed files:

  • Run: git diff --name-only {revision_range}
  • Filter out the spec file itself and any unrelated changes
  • Compare against the "Affected Modules" section in the spec

Get detailed diff:

  • Run: git diff --shortstat {revision_range} to check the size of changes
  • Strategy Selection:
    • Small/Medium Changes (< 300 lines):
      • Run: git diff {revision_range} to get full context in one go
      • Proceed to "Verify Compliance"
    • Large Changes (≥ 300 lines) - Iterative Review Mode:
      • Announce: "This review involves ≥300 lines of changes. Using Iterative Review Mode to process each file individually."
      • Iterate through affected modules:
        • For each module in "Affected Modules", run: git diff {revision_range} -- {module-path}
        • Perform verification checks on this specific chunk
        • Store findings in todos
      • Aggregate all findings at the end

Use TodoWrite to track analysis:

  • Create a todo for each acceptance criterion
  • Create a todo for verifying affected modules
  • Create a todo for running tests

Read the full file on GitHub · 180 lines

Changes

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.

  1. 2d ago First seen · 180 lines · 13 tokens per session scan A 5a4c92bcdf9d

Subscribe to this mod's changes

verify-agent is an agent published in the GitHub repository heiko-braun/draft (5 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 13 tokens to every session and 1,457 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 2026-08-31.

Related

Other agents, from other repositories