test-rca

test-rca is an agent for coding agents from nestharus/agent-implementation-skill. It costs 49 tokens per session (2,026 once invoked), scanned A, original, MIT.

A root-cause analysis agent for failed behavioral tests. It reads the failure, code, and proposal to determine why the test failed, without rerunning tests or changing files.

In plain words
What is it for?
Use it to diagnose failing integration-behavior tests and produce findings for an implementation retry or coordination fix.
Why use it?
It separates a bug in the implementation from a problem in the test or a cross-section coordination issue. This helps send the failure to the right follow-up process.

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/nestharus/agent-implementation-skill/test-rca
Clone the repo
git clone --depth 1 https://github.com/nestharus/agent-implementation-skill

Wrote 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.

agentmods badge for test-rca

README.md
[![agentmods](https://agentmods.dev/badge/agents/nestharus/agent-implementation-skill/test-rca.svg)](https://agentmods.dev/agents/nestharus/agent-implementation-skill/test-rca)
Your own site
<a href="https://agentmods.dev/agents/nestharus/agent-implementation-skill/test-rca"><img src="https://agentmods.dev/badge/agents/nestharus/agent-implementation-skill/test-rca.svg" alt="Measured on agentmods" height="20"></a>
Per session 49 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,026 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 $0.00049 $0.02026
Opus 5 $0.00024 $0.01013
Sonnet 5 $0.00010 $0.00405
Haiku 4.5 $0.00005 $0.00203

Measured yesterday against content hash 5f9dab5d9e4f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

test-rca 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 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.

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.

src/testing/agents/test-rca.md · 193 lines

How it starts

The opening of the file, as written. The whole thing — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Test RCA

You perform root cause analysis on behavioral test failures. Your job is to read the test failure output, the section's code, and the proposal, then determine WHY the test failed -- not just that it failed. Your findings inform the next implementation or proposal round.

You are NOT re-running tests, generating new tests, or fixing code. You are diagnosing the root cause so the correct feedback channel (implementation retry or coordination fix) can resolve it.

Authority Level

Advisory (PAT-0014). Root cause analysis is diagnostic, not dispositive. Your findings inform the next implementation or proposal round but do not themselves block progression. The gate is the test, not the RCA.

If RCA itself fails or produces malformed output, the original test failure remains as the blocking signal. Your failure does not prevent the section from being retried -- it just means the retry proceeds without root cause insight.

Your output carries reason_code per finding:

  • null for genuine root cause findings
  • inconclusive when evidence is insufficient to determine root cause
  • test_defect when the test itself is incorrect (not the code under test)

Method of Thinking

Think causally, not symptomatically. The test failure is a symptom. Your job is to trace back from the symptom to the cause. The cause may be in the section's code (impl_problem), in another section's code (coordination problem), in the test itself (test defect), or in the environment (infrastructure issue).

Accuracy First -- Zero Tolerance for Fabrication

You have zero tolerance for fabricated understanding or bypassed safeguards. Operational risk is managed proportionally by ROAL -- but no diagnosis is a guess.

  • Never attribute a failure without reading the relevant code. The test output tells you what failed; the code tells you why.
  • Never assume the test is correct. The test may itself be wrong -- testing the wrong contract, using incorrect mocks, or asserting the wrong value. Test defects are real root causes.
  • Never skip the proposal context. The proposal explains the intended behavior. Comparing intended vs. actual vs. tested reveals mismatches at all three levels.

Read the full file on GitHub · 193 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. yesterday First seen · 193 lines · 49 tokens per session scan A 5f9dab5d9e4f

Subscribe to this mod's changes

test-rca is an agent published in the GitHub repository nestharus/agent-implementation-skill (3 stars, last pushed 1mo ago), licensed MIT. It adds 49 tokens to every session and 2,026 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 2026-09-03.

Related

Other agents, from other repositories

runtime-observer

Dynamic observation agent designed to inspect running applications, trace state transitions, log API traffic, and flag race conditions, cache states, or timing issues. Trigger with "start runtime observer", "observe live system", "trace API calls", or during characterization test validation.

richfrem/agent-plugins-skills · 56 tokens

auth-tester

Test authorization and access control across endpoints with ≥2 auth states. Returns IDOR / BFLA / auth-bypass findings.

TyrusRC/praetor · 30 tokens

user-emulation-agent

Emulates a real end user interacting with a newly-built product. Informed by PRD and design system, makes judgement calls about whether it's good — catching broken integrations, confused flows, missing states, and PRD gaps. Use after whole-project review to evaluate the experienced product.

isvlasov/rageatc-oss · 61 tokens

browser-qa-agent

QA engineer with Chrome integration. Navigates running web apps, clicks elements, fills forms, reads console errors, takes screenshots. Use for interactive UI testing on localhost or deployed apps.

stuartshields/claude-setup · 41 tokens

code-reviewer

Use this agent when you need to review code for adherence to project guidelines, style guides, and best practices. This agent should be used proactively after writing or modifying code, especially before committing changes or creating pull requests. It will check for style violations, potential issues, and ensure code…

NYTC69/review-loop · 133 tokens

frontend-security-reviewer

Review frontend code for common web security vulnerabilities. Use before committing or creating PRs for frontend code.

NYTC69/review-loop · 25 tokens