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 skills add prime-radiant-inc/greenfield --skill analysis-pipelinegit clone --depth 1 https://github.com/prime-radiant-inc/greenfieldWrote 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/skills/prime-radiant-inc/greenfield/analysis-pipeline)<a href="https://agentmods.dev/skills/prime-radiant-inc/greenfield/analysis-pipeline"><img src="https://agentmods.dev/badge/skills/prime-radiant-inc/greenfield/analysis-pipeline/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/prime-radiant-inc/greenfield/analysis-pipeline"><img src="https://agentmods.dev/badge/skills/prime-radiant-inc/greenfield/analysis-pipeline.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Prompt Injection · line 16 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
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 5.1 | $0.00023 | $0.03562 |
| Opus 5 | $0.00012 | $0.01781 |
| Sonnet 5 | $0.00005 | $0.00712 |
| Haiku 4.5 | $0.00002 | $0.00356 |
Grade A, and why
analysis-pipeline 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 13d 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 — 317 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reverse Engineering
You analyze targets from multiple perspectives. You output behavioral specifications with provenance.
The Core Principle
Analyze deeply. Output behaviorally. Track provenance. Analyze COMPLETELY.
- Analysis: Tear apart the source/binary/runtime. Trace every code path. Understand every decision tree.
- Output: Write specs using behavioral language. No source identifiers in the output.
- Provenance: Every behavioral claim cites its evidence source with
<!-- cite: -->annotations. - Completeness: Analyze ALL modules (P0, P1, P2, P3 - everything). Priority labels are for implementation ordering, NOT for what you analyze.
Exhaustive Reading (CRITICAL)
You MUST read every single line of code. You MUST identify every single routine.
- Do NOT skim code. Do NOT skip "unimportant" sections.
- Do NOT rely on grep patterns alone - READ the actual source.
- Every function, every method, every class, every module must be identified and understood.
- If you haven't read a line, you don't know what it does.
The 7-Layer Pipeline
digraph pipeline {
rankdir=TB;
"Start analysis" [shape=doublecircle];
"Layer 1: Gather intelligence" [shape=box];
"Layer 2: Synthesize and map modules" [shape=box];
"Layer 3: Write deep behavioral specs" [shape=box];
"Gate 1: Verification passed?" [shape=diamond];
"Gate 1b: Source-to-spec completeness" [shape=diamond];
"Layer 4: Generate test vectors and acceptance criteria" [shape=box];
"Gate 2: Spec review passed?" [shape=diamond];
"Layer 5: Sanitize raw specs to output/" [shape=box];
"Layer 6: Second-pass review passed?" [shape=diamond];
"Layer 7: Fidelity validation passed?" [shape=diamond];
"Analysis complete" [shape=doublecircle];
"Remediate Gate 1 findings" [shape=box];
"Remediate Gate 2 findings" [shape=box];
"Remediate Layer 6 findings" [shape=box];
"Remediate Layer 7 findings" [shape=box];
"STOP: Gate 1 failed after 3 attempts" [shape=octagon, style=filled, fillcolor=red, fontcolor=white];
"STOP: Gate 2 failed after 3 attempts" [shape=octagon, style=filled, fillcolor=red, fontcolor=white];
"STOP: Layer 6 failed after 3 attempts" [shape=octagon, style=filled, fillcolor=red, fontcolor=white];
"STOP: Layer 7 failed after 3 attempts" [shape=octagon, style=filled, fillcolor=red, fontcolor=white];
"Start analysis" -> "Layer 1: Gather intelligence";
"Layer 1: Gather intelligence" -> "Layer 2: Synthesize and map modules";
"Layer 2: Synthesize and map modules" -> "Layer 3: Write deep behavioral specs";
"Layer 3: Write deep behavioral specs" -> "Gate 1: Verification passed?";
"Gate 1: Verification passed?" -> "Gate 1b: Source-to-spec completeness" [label="yes"];
"Gate 1: Verification passed?" -> "Remediate Gate 1 findings" [label="no"];
"Remediate Gate 1 findings" -> "Gate 1: Verification passed?" [label="attempt < 3"];
"Remediate Gate 1 findings" -> "STOP: Gate 1 failed after 3 attempts" [label="attempt >= 3"];
"Gate 1b: Source-to-spec completeness" -> "Layer 4: Generate test vectors and acceptance criteria" [label="pass"];
"Gate 1b: Source-to-spec completeness" -> "Remediate Gate 1 findings" [label="gaps found"];
"Layer 4: Generate test vectors and acceptance criteria" -> "Gate 2: Spec review passed?";
"Gate 2: Spec review passed?" -> "Layer 5: Sanitize raw specs to output/" [label="yes"];
"Gate 2: Spec review passed?" -> "Remediate Gate 2 findings" [label="no"];
"Remediate Gate 2 findings" -> "Gate 2: Spec review passed?" [label="attempt < 3"];
"Remediate Gate 2 findings" -> "STOP: Gate 2 failed after 3 attempts" [label="attempt >= 3"];
"Layer 5: Sanitize raw specs to output/" -> "Layer 6: Second-pass review passed?";
"Layer 6: Second-pass review passed?" -> "Layer 7: Fidelity validation passed?" [label="yes"];
"Layer 6: Second-pass review passed?" -> "Remediate Layer 6 findings" [label="no"];
"Remediate Layer 6 findings" -> "Layer 6: Second-pass review passed?" [label="attempt < 3"];
"Remediate Layer 6 findings" -> "STOP: Layer 6 failed after 3 attempts" [label="attempt >= 3"];
"Layer 7: Fidelity validation passed?" -> "Analysis complete" [label="yes"];
"Layer 7: Fidelity validation passed?" -> "Remediate Layer 7 findings" [label="no"];
"Remediate Layer 7 findings" -> "Layer 7: Fidelity validation passed?" [label="attempt < 3"];
"Remediate Layer 7 findings" -> "STOP: Layer 7 failed after 3 attempts" [label="attempt >= 3"];
}
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.
- 13d ago First seen · 317 lines · 23 tokens per session scan A c90dae90262f
analysis-pipeline is a skill published in the GitHub repository prime-radiant-inc/greenfield (276 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 23 tokens to every session and 3,562 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 skills, from other repositories
benchmark
Local-only regression / benchmark skill for ui-clone-skills maintainers. Drives the standard ui-reverse-engineering pipeline against the canonical reference site (https://realfood.gov) and records AE/SSIM, iteration count, gate fail counts, and outcome to benchmark/history.csv so prompt / sub-doc / model-version drift…
write-test-plans
Interactive workflow for writing new Speky test plans for requirements lacking coverage, guided by the Speky MCP server.
speky-workflow
Guidelines for working with Speky requirements and tests via the speky MCP server tools.
SPI Flash Dump via OpenOCD
This skill should be used when the user asks to "dump SPI flash", "read EEPROM through OpenOCD", "dump flash via MCU", "read SPI through debug interface", "extract firmware from SPI", or mentions scenarios involving SPI flash connected to a microcontroller with SWD/JTAG debug access. Provides comprehensive guidance…
create-site
Creates a new Power Pages code site (SPA) using React, Angular, Vue, or Astro. Guides through the full process from initial concept to deployed site: requirements discovery, scaffolding, component planning, design, implementation, validation, and deployment. Use when the user wants to create, build, or scaffold a new…
review
5-pass structured code review — correctness, security, performance, readability, consistency.