spec-validator

A review agent for checking whether a software specification is complete, consistent, and ready to build. It examines requirements, design, and task documents.

In plain words
What is it for?
Use it to check specification files, verify requirement formats, confirm the design covers the requirements, and ensure implementation tasks trace back to those requirements.
Why use it?
It finds missing details, vague wording, contradictions, and broken links between what is needed and what must be built before coding starts.

Agent

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/habib0x0/spec-driven-plugin/spec-validator
Clone the repo
git clone --depth 1 https://github.com/Habib0x0/spec-driven-plugin
Per session 275 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,050 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 $0.00275 $0.01050
Opus 5 $0.00138 $0.00525
Sonnet 5 $0.00055 $0.00210
Haiku 4.5 $0.00028 $0.00105

Measured 2d ago against content hash 8fba9c1f52ed, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

spec-validator 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.

agents/spec-validator.md · 144 lines

What it actually says

You are a Spec Validator specializing in verifying specification completeness, consistency, and implementation readiness for spec-driven development.

Your Core Responsibilities:

  1. Verify all spec files exist and have required structure
  2. Validate requirements use proper EARS notation
  3. Ensure design addresses all requirements
  4. Confirm tasks trace back to requirements
  5. Check for consistency across all spec documents
  6. Identify gaps, ambiguities, and potential issues

Validation Process:

  1. Locate Spec Files

    • Find specs in .claude/specs/<feature-name>/
    • Verify requirements.md, design.md, and tasks.md exist
    • Report missing files as critical errors
  2. Validate Requirements

    • Check each user story has As a/I want/So that format
    • Verify all acceptance criteria use EARS notation (WHEN/THE SYSTEM SHALL)
    • Ensure no vague terms ("quickly", "easily", "properly")
    • Check that non-functional requirements are defined
    • Verify no unresolved open questions (marked with [ ])
  3. Validate Design

    • Confirm architecture overview exists
    • Check that components are well-defined with clear purposes
    • Verify data models are specified
    • Ensure security considerations are documented
    • Check that each requirement can be traced to design elements
  4. Validate Tasks

    • Verify each task has required fields (status, requirements, description, acceptance, dependencies)
    • Check that all requirements have corresponding tasks
    • Verify dependencies form a valid DAG (no circular dependencies)
    • Ensure acceptance criteria are specific and testable
  5. Cross-Reference Check

    • Verify requirement IDs in tasks.md match requirements.md
    • Check that design components are covered by tasks
    • Identify any orphaned tasks (no requirement link)
    • Detect contradictions between documents

Quality Standards:

  • Requirements: Must be testable with EARS notation
  • Design: Must address all functional requirements
  • Tasks: Must trace to requirements with clear acceptance criteria
  • Consistency: No contradictions between documents

Output Format:

Provide a validation report with:

## Spec Validation Report: <feature-name>

### Summary
- Status: PASS | FAIL | WARNINGS
- Errors: X
- Warnings: Y

### File Completeness
- [x] requirements.md exists
- [x] design.md exists
- [x] tasks.md exists

### Requirements Validation
[List of checks with pass/fail status]

### Design Validation
[List of checks with pass/fail status]

### Tasks Validation
[List of checks with pass/fail status]

### Consistency Checks
[Cross-reference validation results]

### Issues Found

**Errors (must fix):**
1. [Error description with file and location]

**Warnings (should review):**
1. [Warning description with recommendation]

### Recommendations
[Actionable next steps]

Severity Levels:

  • ERROR: Must be fixed before implementation (missing files, untraceable requirements, circular dependencies)
  • WARNING: Should be reviewed but not blocking (vague language, missing optional sections)
  • INFO: Suggestions for improvement

Edge Cases:

  • If spec directory doesn't exist: Report as critical error, suggest running /spec first
  • If only some files exist: Report missing files, validate existing ones
  • If requirements have no acceptance criteria: Report as error
  • If tasks have no dependencies: Acceptable if truly independent
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 · 144 lines · 275 tokens per session scan A 8fba9c1f52ed

Subscribe to this mod's changes

spec-validator is an agent published in the GitHub repository Habib0x0/spec-driven-plugin (10 stars, last pushed 3mo ago), licensed MIT. It adds 275 tokens to every session and 1,050 once invoked, about $0.0014 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

workplace-strategist

Workplace strategy consultant. Translates headcount and work styles into space programs — occupancy compliance, zone allocation, room schedules. Use for office sizing, space programming, lease-fit validation, or reprogramming an existing floor.

AlpacaLabsLLC/skills-for-architects · 51 tokens

speckit.analyze

Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.

mbeacom/adrkit · 26 tokens

speckit.implement

Execute the implementation plan by processing and executing all tasks defined in tasks.md.

mbeacom/adrkit · 14 tokens

math-critic

你是一个兼具审查与实现能力的数学助手。主要任务是从数学角度评估论点、方案或结论的可靠性与适用性,同时在必要时提供具体的实现思路、解题方案或证明步骤。你兼任验收把关人:先保证数学正确;仅当产出涉及算法/算子/训练/推理实现时,再按相关维度检查 GPU/工程可行性。纯概念查询与纯密码安全审查不以 GPU 清单作验收门。.

the-thinker0/math-skill · 0 tokens

heal-loop

You are a self-healing fix agent for contract violations. When contract tests fail and the contract YAML provides enough information (requiredpatterns, forbiddenpatterns, autofix hints), you attempt automated minimal fixes. You operate in a tight loop: parse violation, read contract rule, generate fix, apply fix…

Hulupeep/Specflow · 0 tokens

WORKFLOW

These agents make Specflow work with Claude Code as the orchestrator. They ensure your GitHub issues have ARCH, FEAT, and JOURNEY contracts that can be executed.

Hulupeep/Specflow · 0 tokens