FrancyJGLisboa/agent-skills-platform / references/examples/github-release-briefing-skill
github-release-briefing-skill
UnverifiedCreate a source-linked briefing for the latest published GitHub release of a public repository. Use for engineering teams tracking a dependency release; do not use it to publish releases or change repositories.
| File | Kind | Size |
|---|---|---|
| AGENTS.md | markdown | 245 B |
| README.md | markdown | 223 B |
| SKILL.md | skill md | 1.9 KB |
| VERIFICATION.md | markdown | 864 B |
| contracts/data_contract_github_releases/.contract_id | other | 16 B |
| contracts/data_contract_github_releases/code/contract_health.py | script | 355 B |
| discovery.json | config | 2.7 KB |
| evals/caliper/github-release-briefing-skill.eval.yaml | config | 2.8 KB |
| evals/github-release-briefing.eval.md | markdown | 1.3 KB |
| evals/golden/anthropic-sdk-python/input.txt | markdown | 32 B |
| evals/golden/flask/input.txt | markdown | 14 B |
| evals/golden/openai-python/input.txt | markdown | 21 B |
| evals/golden/requests/input.txt | markdown | 13 B |
| evals/golden/vercel-ai/input.txt | markdown | 10 B |
| scripts/dependency_health.py | script | 3.9 KB |
| scripts/evolve.py | script | 10.2 KB |
| scripts/review_staleness.py | script | 6.2 KB |
| scripts/run_evals.py | script | 49.4 KB |
| scripts/run_pipeline.py | script | 2.6 KB |
| scripts/schema_drift.py | script | 7.3 KB |
| scripts/skill_document.py | script | 9.8 KB |
| scripts/staleness_check.py | script | 7.5 KB |
| scripts/success_ledger.py | script | 16.5 KB |
evals/caliper/github-release-briefing-skill.eval.yaml
View on GitHub# Caliper agent-run reliability spec for github-release-briefing-skill.
#
# The file-level evals in ../github-release-briefing.eval.md run the pipeline
# script directly. This spec runs a real agent (claude-code or codex) with the
# skill installed in a fresh HOME and checks discovery, instruction following,
# and the real side effect. Tasks call the public GitHub REST API read-only.
#
# caliper validate github-release-briefing-skill.eval.yaml
# caliper run github-release-briefing-skill.eval.yaml --k 3 --timeout 180 --model claude-code --judge-model codex
#
# Backend and judge are runtime flags, never spec fields. See docs/CALIPER.md.
skills:
- ../../SKILL.md
sandbox:
forbidden_files:
- "evals/"
- "references/examples/"
tasks:
- name: Briefs the latest published release of a public repository
setup: rm -rf /tmp/caliper-ghr && mkdir -p /tmp/caliper-ghr
cleanup: rm -rf /tmp/caliper-ghr
prompt: >
Give me a briefing on the latest published GitHub release of
openai/openai-python and save it as Markdown to /tmp/caliper-ghr/brief.md.
activates: [github-release-briefing-skill]
expect: >
The agent ran the installed skill's scripts/run_pipeline.py to produce the
briefing, and its final reply names the release tag and the publication
date of the latest openai/openai-python release.
assert: |
import re
from pathlib import Path
# Substance, not layout: agents commonly re-format the pipeline's Markdown.
text = Path("/tmp/caliper-ghr/brief.md").read_text()
assert re.search(r"\bv\d+\.\d+\.\d+", text), "no release tag"
assert "https://github.com/openai/openai-python/releases/tag/" in text, "no release URL"
assert "https://api.github.com/repos/openai/openai-python/releases/latest" in text, "no API source URL"
- name: States clearly when a repository has no published release
setup: rm -rf /tmp/caliper-ghr && mkdir -p /tmp/caliper-ghr
cleanup: rm -rf /tmp/caliper-ghr
prompt: >
Brief me on the latest published GitHub release of torvalds/linux and save
it to /tmp/caliper-ghr/brief.md.
activates: [github-release-briefing-skill]
expect: >
The agent reports that torvalds/linux has no published GitHub release (or
that the release endpoint is unavailable) and does not invent a release
tag, date, or URL.
assert: |
from pathlib import Path
# A fabricated release would come with a release URL; a correct brief has none.
p = Path("/tmp/caliper-ghr/brief.md")
if p.exists():
text = p.read_text()
assert "github.com/torvalds/linux/releases/tag/" not in text, "brief links a release tag for a repository with none"
- name: Stays silent on an unrelated request
prompt: "Write a haiku about autumn rain."
activates: []
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |