A skill that provides programmatic access to GitHub for the hallucination-detector repository. It uses the GitHub API through Node.js scripts rather than requiring the GitHub command-line tool.

Use it to work with issues, pull requests, reviews, labels, milestones, Projects, CI workflow runs, check-run annotations, and logs through GitHub’s API.

It lets agents manage repository work in environments where the GitHub command-line tool or direct GitHub network access is unavailable. It also handles proxy settings automatically.

gh — Skill by bitflight-devops

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/bitflight-devops/hallucination-detector/gh

Any agent

npx skills add bitflight-devops/hallucination-detector --skill gh

Clone the repo

git clone --depth 1 https://github.com/bitflight-devops/hallucination-detector

Made for: Claude Code, Codex.

Per session 49 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,859 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.

Model	Per session	Once invoked
Fable 5	$0.00049	$0.01859
Opus 5	$0.00024	$0.00929
Sonnet 5	$0.00010	$0.00372
Haiku 4.5	$0.00005	$0.00186

Measured 2d ago against content hash 9e3d91c9564d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

gh 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 2d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/github-project-setup.cjs), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

.claude/skills/gh/SKILL.md · 182 lines

How it starts

The opening of the file, as written. The whole thing — 182 lines — stays where its author put it; the contents beside it link to each section on GitHub.

GitHub API — Setup and Usage

Purpose

Provides programmatic GitHub API access for the hallucination-detector repo using octokit (full SDK) with automatic proxy support. All operations use Node.js scripts — no gh CLI binary is needed.

When to Use

Need to interact with GitHub API (issues, PRs, labels, milestones)
Creating or listing pull requests
Reviewing PRs: list/view/submit/dismiss reviews, code annotation comments
Managing CI: list/view/rerun workflow runs, check run annotations and logs
Managing GitHub Issues, Projects V2, or Labels
Running in a proxy environment where direct DNS to api.github.com is unavailable

Authentication

GITHUB_TOKEN environment variable provides authentication. All scripts read it automatically via createGitHubClient().

const { createGitHubClient, OWNER, REPO } = require('./.claude/scripts/lib/github-client.cjs');
const octokit = createGitHubClient();

The client auto-detects HTTPS_PROXY / HTTP_PROXY env vars and routes requests through the egress proxy via undici.ProxyAgent. No manual proxy configuration needed.

Scripts

General-purpose GitHub API — `gh-api.cjs`

# Issues
node .claude/scripts/gh-api.cjs issue list
node .claude/scripts/gh-api.cjs issue create --title "feat: add X" --label "type:feature" --label "priority:p1"
node .claude/scripts/gh-api.cjs issue view 42
node .claude/scripts/gh-api.cjs issue comment 42 --body "Implemented in PR #45."
node .claude/scripts/gh-api.cjs issue comment list 27 [--user "coderabbitai[bot]"]
node .claude/scripts/gh-api.cjs issue comment view 123456
node .claude/scripts/gh-api.cjs issue comment search 27 --user "coderabbitai[bot]" --section "Prompt for AI Agents" --source reviews

# Pull Requests
node .claude/scripts/gh-api.cjs pr list
node .claude/scripts/gh-api.cjs pr create --title "feat: add X" --base main --body "Details"

# Labels
node .claude/scripts/gh-api.cjs label list
node .claude/scripts/gh-api.cjs label create --name "priority:p0" --color "D73A4A" --description "Critical"

# PR Reviews
node .claude/scripts/gh-api.cjs review list 27
node .claude/scripts/gh-api.cjs review view 27 123456
node .claude/scripts/gh-api.cjs review submit 27 --event APPROVE --body "LGTM"
node .claude/scripts/gh-api.cjs review dismiss 27 123456 --message "Outdated"

# Code Annotation Comments (PR review comments)
node .claude/scripts/gh-api.cjs review-comment list 27
node .claude/scripts/gh-api.cjs review-comment view 123456
node .claude/scripts/gh-api.cjs review-comment reply 27 123456 --body "Fixed"

# CI Check Runs (accepts PR number or git ref/SHA)
node .claude/scripts/gh-api.cjs checks list 27
node .claude/scripts/gh-api.cjs checks list main
node .claude/scripts/gh-api.cjs checks view 123456
node .claude/scripts/gh-api.cjs checks annotations 123456

# Workflow Runs
node .claude/scripts/gh-api.cjs run list [--limit 10] [--status completed]
node .claude/scripts/gh-api.cjs run view 123456
node .claude/scripts/gh-api.cjs run rerun 123456 [--failed-only]
node .claude/scripts/gh-api.cjs run logs 123456
node .claude/scripts/gh-api.cjs run cancel 123456

Read the full file on GitHub · 182 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

2d ago First seen · 182 lines · 49 tokens per session scan A 9e3d91c9564d

Subscribe to this mod's changes

gh is a skill published in the GitHub repository bitflight-devops/hallucination-detector (7 stars, last pushed 29d ago), licensed MIT. It adds 49 tokens to every session and 1,859 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 2026-08-31.

AI Safety Git and workflow Hallucination Detection LLM-guardrails Project management claude-code claude-code-plugin developer-tools

implementing-llm-guardrails-for-security

Implements input and output validation guardrails for LLM-powered applications to prevent prompt injection, data leakage, toxic content generation, and hallucinated outputs. Builds a security validation pipeline using NVIDIA NeMo Guardrails Colang definitions, custom Python validators for PII detection and content…

xalgorix/xalgorix · 143 tokens

drive

Drive the Polis agent pipeline end-to-end from the terminal — list issues/PRs and their stage, create issues, apply any agent: trigger label (plan / spec / code / fix / review), watch the Actions run, read review feedback, and merge — without touching the GitHub web UI. Use when the user wants to run, drive, advance…

permit0-ai/permit0 · 81 tokens

handing-off-work

Hands off unfinished work with a closed-loop briefing of state, changed conditions, remaining scope, authority limits, and open evidence. Use when AI-agent, reviewer, verifier, releaser, or resumed-thread work transfers to a new owner. Do not use when the same owner continues uninterrupted with full context.

FlyFission/nuclear-grade-context-engineering · 65 tokens

roadmap-sync

Phase 10.6 — sync the project's execution roadmap (project.json → roadmap.path, default docs/roadmap-execution-plan.md) to the just-landed work. Flips the tasks named in workflow.json → roadmaptasks[] from ⬜ to ✅ and promotes their epic headings ⬜→🟡→✅, preserving the standup parser format contract. On the epic track…

friedbotstudio/baseline · 171 tokens

retrospective

Cycle-end retrospective converting recurring judgment failures into durable enforcement. Reviews the cycle's friction — repeated user corrections, guard trips, RCA themes, integrate failures — and turns each recurring one into a landmines.md entry, then proposes graduation candidates up the enforcement funnel…

friedbotstudio/baseline · 139 tokens

power

Power batch-sprint helper — hosts the two behaviours that distinguish the power workflow track from spec-entry. (1) Per-ticket iteration: security runs once PER TICKET over workflow.json.tickets[] while the mechanical phases run once for the batch. (2) Commit split: at the commit phase, group the batch's working tree…

friedbotstudio/baseline · 112 tokens