claude-code-plugin-ref…
Explain plugin, skill, command, agent, and hook mechanics used here. Use when authoring or debugging plugins. Do not use for ops; use night-market-operations.
Audits the DSA problem bank for coverage gaps and proposes new YAML entries. Use when refreshing the problem bank during update-plugins runs.
$ npx -y skills add athola/claude-night-market --skill gauntlet-curate --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/gauntlet-curateContext preview
The summary Claude sees to decide when to auto-load this skill.
Audits the DSA problem bank for coverage gaps and proposes new YAML entries. Use when refreshing the problem bank during update-plugins runs.
name: gauntlet-curate description: Audits the DSA problem bank for coverage gaps and proposes new YAML entries. Use when refreshing the problem bank during update-plugins runs. model_hint: standard
Survey the DSA problem bank, identify coverage gaps, and propose new YAML entries for human review.
`gauntlet:curate`)
Invoke this skill manually with `Skill(gauntlet:gauntlet-curate)` when the problem bank needs a coverage review. The skill is intended to participate in `/update-plugins` runs but is not yet wired into that command (see openpackage.yml registration). It is distinct from `gauntlet:curate`, which handles per-annotation knowledge capture and is what `/gauntlet-curate` invokes today.
1. **Locate the problem bank** at `plugins/gauntlet/data/problems/`. Read `_manifest.yaml` to load the expected NeetCode counts per category.
2. **Survey current coverage** by counting problems in each YAML file (skipping `_manifest.yaml`). Run the analysis script:
cd plugins/gauntlet python scripts/curate_problems.py data/problems/ --output /tmp/gauntlet-curate-report.md
3. **Identify gaps**: categories whose actual count falls below the `neetcode_count` in the manifest. The script sorts gaps largest-first so the worst shortfalls appear at the top.
4. **Review existing problems** in each gap category to understand what is already covered before proposing additions.
5. **Propose new YAML entries** following the schema below. Add proposals to the report under "Proposed New Problems". Do NOT write proposals directly into `data/problems/*.yaml`.
6. **Validate proposals** by running:
python -c "
import yaml, sys
sys.path.insert(0, 'src')
from gauntlet.models import BankProblem
proposals = yaml.safe_load(open('proposals.yaml'))
for p in proposals:
BankProblem.from_dict(p)
print('All proposals valid.')
"7. **Present the report** to the human for review. The report includes the coverage table, gap list, and proposed entries. The human decides which proposals to merge into the YAML files.
Each proposed entry must follow this schema:
- id: category-NNN
title: Problem Title
difficulty: easy # easy | medium | hard | extra_hard
prompt: |
Problem statement with constraints and examples.
hints:
- First hint.
- Second hint.
solution_outline: |
Approach and time/space complexity.
tags: [tag1, tag2]
neetcode_id: neetcode-NNN
challenge_type: explain_why # explain_why | multiple_choice | trace
# | code_complete | debug | rankRequired fields: `id`, `title`, `difficulty`, `prompt`. Optional fields default to empty values.
intentionally has none.
A markdown report at the path specified by `--output`, containing:
Human review is required before any YAML file changes.
path given to `--output` (e.g., `/tmp/gauntlet-curate-report.md`)
`neetcode_count` versus actual count for every category in `_manifest.yaml`, with gaps sorted largest-first
all proposals appear only in the report under "Proposed New Problems"
validation (required fields: `id`, `title`, `difficulty`, `prompt`) before appearing in the report
A plugin marketplace for Claude Code. Install only the plugins you need to run git workflows, code review, spec-driven development, and autonomous agents from inside your Claude Code session.
Explain plugin, skill, command, agent, and hook mechanics used here. Use when authoring or debugging plugins. Do not use for ops; use night-market-operations.
States load-bearing decisions, invariants, and weak points. Use when judging a design change. Do not use for gating; use night-market-change-control.
Rebuild the dev environment: uv, Python tiers, pins, traps. Use when onboarding or toolchain breaks. Do not use for daily commands; use night-market-operations.
Classify, gate, and review changes. Use when landing a PR, releasing, or amending rules. Do not use for failure triage; use night-market-debugging-playbook.
Search and record project memory (Discussions, journal, ADRs). Use before re-investigating anything. Do not use for settled battles; see failure-archaeology.
Bind loop 'done' to unfakeable gates. Use to harden egregore/herald loops or promote completion_integrity. Not for QA gates; use night-market-validation-and-qa.