Nm Gauntlet Challenge

Run a gauntlet challenge session with adaptive difficulty. Tests codebase understanding through multiple choice, code completion, trace exercises, and more

Security Scan

The skill's requested actions (reading/writing files under .gauntlet and evaluating answers) match its description and there are no unexpected installs, credentials, or network behaviors declared.

Purpose & Capability

Name/description describe an in-repo gauntlet that quizzes developers. The SKILL.md only requires access to repository-local state files (.gauntlet/knowledge.json, .gauntlet/state/*) and updating developer progress — these needs are coherent with the stated purpose.

Instruction Scope

Instructions explicitly tell the agent to read .gauntlet/knowledge.json and developer progress files, check pending_challenge.json, evaluate answers, and write a pass token for pre-commit gates. Reading and writing repository files is in-scope for a codebase quiz, but users should be aware this skill will access and modify files in the repo (the skill does not limit which files beyond the .gauntlet paths). The evaluation step refers to the "developer's most recent message," which may cause the agent to consult conversation context; the SKILL.md is not prescriptive about exactly which contexts to use.

Install Mechanism

No install spec and no code files — this is instruction-only, so nothing is written to disk by an installer. Low installation risk.

Credentials

The skill declares no required environment variables, credentials, or external service keys. Its file-access requirements are proportional to a local gauntlet that stores knowledge and state in .gauntlet.

Persistence & Privilege

The skill is not marked always:true and does not request elevated platform privileges. However, it does perform writes to repository state (e.g., pass tokens, progress/streak updates). That is normal for this use case but means the skill can modify repository files when invoked.

Assessment

This skill appears to do exactly what it says: read and update files under a .gauntlet directory to run quizzes. Before enabling it, inspect the .gauntlet/knowledge.json and state files in your repo to ensure they contain no secrets and that you trust the pass-token behavior (it will write tokens to the repo when gating). If you need stronger isolation, run the skill on a disposable clone or ensure your CI/pre-commit setup verifies pass tokens. Note the SKILL.md is somewhat high-level about how answers are evaluated and which conversation context is consulted — if you need strict auditability, test the skill in a controlled environment first.

Night Market Skill — ported from claude-night-market/gauntlet. For the full experience with agents, hooks, and commands, install the Claude Code plugin.

Run Gauntlet Challenge

Present challenges from the knowledge base and evaluate answers.

Steps

1. Load state: read .gauntlet/knowledge.json and developer progress

2. Check for pending challenge: if .gauntlet/state/pending_challenge.json exists, evaluate the developer's most recent message as an answer before generating a new one

3. Generate challenge: use adaptive weighting to select a knowledge entry and challenge type

4. Present challenge: show the question with context

5. Evaluate answer: score the response (pass/partial/fail)

6. Record result: update developer progress and streak

7. On pass: write pass token if from pre-commit gate. Show next challenge if in session.

8. On fail: show correct answer with explanation. Present a new challenge.

Scoring

Result	Score	Streak
Pass	1.0	+1
Partial	0.5	reset
Fail	0.0	reset