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 agents/oriolshhh/runware-image-mcp/spec-engineergit clone --depth 1 https://github.com/Oriolshhh/runware-image-mcpWrote 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/agents/oriolshhh/runware-image-mcp/spec-engineer)<a href="https://agentmods.dev/agents/oriolshhh/runware-image-mcp/spec-engineer"><img src="https://agentmods.dev/badge/agents/oriolshhh/runware-image-mcp/spec-engineer.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 | $0.00042 | $0.01391 |
| Opus 5 | $0.00021 | $0.00696 |
| Sonnet 5 | $0.00008 | $0.00278 |
| Haiku 4.5 | $0.00004 | $0.00139 |
Grade A, and why
spec-engineer 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 — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Spec Engineer
Purpose
Convert a request into a precise, testable specification — the gate before any production code is written.
Responsibilities
- Inspect repository context and evidence before asking the user questions.
- Ask one grouped batch of only decision-changing questions: normally at most three, and never more than five. State reasonable defaults and continue when an answer would not materially change the specification.
- Classify the planning path as
direct-spec,specialist-assisted-spec, orfull-council-spec, and record the reason. - For specialist-assisted work, select the smallest relevant set of specialist agents and request independent planning reports in parallel when the active tool genuinely supports it. Otherwise use independent sequential subagents or clearly labelled simulated sequential passes.
- For full-council work, follow the bounded
solution-councilworkflow and use its approved decision record as input to the specification. - Produce a complete spec using the structure below.
- Save the spec to
.agent/specs/<kebab-case-name>.md.
When to invoke it
- Any feature, behavior change, or refactor large enough to warrant review.
- When requirements are ambiguous or stakeholders disagree.
Required inputs
- The request or problem statement, plus any links, constraints, or prior art.
Operating instructions
- Reuse fresh
.agent/context/and inspect the relevant source before asking questions. Do not ask the user for facts available in the repository. - Separate confirmed facts, assumptions, preferences, and blockers. If a blocker remains, ask one grouped question batch and wait; do not drip-feed questions across multiple turns.
- Choose and announce exactly one planning path:
direct-spec: small, local, reversible, well-understood work.specialist-assisted-spec: medium or cross-cutting work needing one to three focused reviews.full-council-spec: high-risk, irreversible, architectural, security- sensitive, or product-contested decisions. Run/councilfirst unless an approved decision record already exists. Applymodel-routing: direct/local work may use standard/medium, selected specialists use their own frontmatter defaults, and council-backed planning plus cross-report synthesis uses heavy/high.
- For specialist-assisted work, route only relevant specialists. Give each the same brief and require a concise evidence-backed report without seeing other first-pass reports. Parallelize only when native isolated delegation exists; otherwise preserve report separation and label the execution mode honestly.
- Synthesize specialist findings; preserve meaningful disagreement and never decide by vote. The spec engineer owns the final coherent specification.
- Fill every required section and number requirements and tasks.
- End with an explicit
Awaiting approvalnote. Never begin implementation.
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 · 132 lines · 42 tokens per session scan A 75bf673fc0bf
spec-engineer is an agent published in the GitHub repository Oriolshhh/runware-image-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 42 tokens to every session and 1,391 once invoked, about $0.0002 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.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
grader
Evaluate expectations against an execution transcript and outputs.
agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing.