qcoder-self-evaluator

A testing and security-review guide for QCoder changes, prompts, agents, and authorized applications. It combines repeatable tests with controlled checks for unsafe behavior, prompt attacks, data leaks, and authorization problems.

In plain words
What is it for?
Use it for regression gates, adversarial prompt testing, model-safety checks, agent tool-choice tests, and limited application-security retests.
Why use it?
It helps teams compare a change with a recorded baseline and catch regressions before release. It also requires authorization and scoped testing before sending active security checks to an application.

Skill for Claude CodeCodex

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 skills/echoomegaprime/echo-qcoder/qcoder-self-evaluator
Any agent
npx skills add echoomegaprime/echo-qcoder --skill qcoder-self-evaluator
Clone the repo
git clone --depth 1 https://github.com/echoomegaprime/echo-qcoder

Made for: Claude Code, Codex.

Per session 85 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 461 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 $0.00085 $0.00461
Opus 5 $0.00043 $0.00230
Sonnet 5 $0.00017 $0.00092
Haiku 4.5 $0.00009 $0.00046

Measured yesterday against content hash 8ac33667017f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

qcoder-self-evaluator 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.

qwen-skills/qcoder-self-evaluator/SKILL.md · 33 lines

How it starts

The opening of the file, as written. The whole thing — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.

QCoder Self Evaluator

Expected input

Define the immutable candidate, baseline, acceptance thresholds, authorized target when applicable, permitted probe families, time and request budgets, and artifact destination.

Evaluation sequence

  1. Freeze candidate identity and record environment, model route, code revision, dataset revision, and configuration.
  2. Run deterministic unit, integration, protocol, security, and packaging checks before model-judged evaluations.
  3. Use the repository's labelled golden prompts for activation, tool choice, arguments, false positives, false negatives, and prohibited behavior.
  4. Use QCoder's local golden evaluator for repeatable model or agent comparisons. Promptfoo remains a pinned reference but is withheld from installation while its current dependency audit contains reachable high-severity findings; never weaken the audit gate to install it.
  5. Route Garak model probes and Nuclei application validation through isolated, scoped sidecars. For Nuclei or any active target traffic, switch to the pentester role and validate authorization first.
  6. Compare against the recorded baseline. Fail closed on unsafe activation, secret leakage, authorization bypass, destructive mislabelling, or a required deterministic gate.
  7. Save redacted machine-readable results and a human summary. A model grader may supplement but never replace executable evidence.

Recovery

Classify failures as product defect, test defect, environment defect, upstream drift, or external blocker. Repair only the owning layer, rerun the failed gate, then rerun the full ladder before release.

Stop conditions

Stop on target-scope mismatch, secret-bearing untrusted evaluation content, uncontrolled remote generation, missing immutable identity, resource exhaustion, or a doctrine hard limit.

Output format

Report candidate and baseline IDs, gates and counts, adversarial results, regressions, residual risk, artifact hashes, and promotion or rejection decision.

Read the full file on GitHub · 33 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 · 33 lines · 85 tokens per session scan A 8ac33667017f

Subscribe to this mod's changes

qcoder-self-evaluator is a skill published in the GitHub repository echoomegaprime/echo-qcoder (0 stars, last pushed yesterday), licensed MIT. It adds 85 tokens to every session and 461 once invoked, about $0.0004 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 skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens