newsroom-harness
Skill397 starsChanged 10 days agoApache-2.0View on GitHub
Slip into the Newsroom agent harness — combines an investigative reporter persona (OCeaN), traffic-light permissions, graph memory with dreaming, local reporting skills, and three deterministic referees (Truth, Safety, and Reading grade).
In this repository79 other files it gives its agents.
- agents/newsroom-harness/AGENTS.mdAGENTS.md
- codelabs/ai-toolkit-lab-2/hackathon-judge/frontend/.agents/skills/vercel-composition-patterns/AGENTS.mdAGENTS.md
- codelabs/ai-toolkit-lab-2/hackathon-judge/frontend/.agents/skills/vercel-react-best-practices/AGENTS.mdAGENTS.md
- codelabs/bwg002-devrel-2/AGENTS.mdAGENTS.md
- calendar-exportSkill
- session-notesSkill
- calendar-exportSkill
- session-notesSkill
- calendar-exportSkill
- session-notesSkill
- calendar-exportSkill
- session-notesSkill
- compare-subagent-mcp-visibilitySkill
- batch-executing-test-suitesSkill
- circuit-breaker-reporterSkill
- generating-mock-imagesSkill
- making-forms-demoSkill
- multi-step-workflow-reporterSkill
- auditing-inventorySkill
- verify-sourcesSkill
- write-bulletinSkill
- repo-walkthroughSkill
- go-project-setupSkill
- latest-software-versionSkill
- video-descriptionSkill
- adk-memory-bank-initializerSkill
- agent-containerizerSkill
- cloud-run-agent-architectSkill
- gcp-agent-eval-engine-runnerSkill
- gcp-agent-eval-metric-configuratorSkill
- gcp-agent-golden-dataset-builderSkill
- gcp-agent-model-armor-shieldSkill
- gcp-agent-safety-gatekeeperSkill
- gcp-agent-sdp-template-factorySkill
- gcp-agent-shadow-deployerSkill
- gcp-agent-tool-trajectory-evaluatorSkill
- gcp-production-secret-handlerSkill
- mcp-connector-generatorSkill
- go-architectSkill
- go-backend-devSkill
- go-reviewerSkill
- go-test-expertSkill
- vercel-composition-patternsSkill
- vercel-react-best-practicesSkill
- lab-helperSkill
- start-frontendSkill
- brand-guidelinesSkill
- brand-guidelinesSkill
- knowledge_catalog_governanceSkill
- banana-breadSkill
- cheese-breadSkill
- drawingsSkill
- paintingsSkill
- photographySkill
- ticket-triageSkill
- adding-easter-eggSkill
- building-platformer-gamesSkill
- creating-a-gameSkill
- ensuring-arcade-visualsSkill
- handling-user-inputSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 2 are distinctive, and 6 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 9
Use this skill whenever the user asks you to "slip into the Newsroom harness", investigate a breaking news story, fact-check a draft, or …
- Very specific: line 14
Read identity/persona.md, identity/job.md, identity/style.md, and permissions.md.
- Very specific: line 16
Check memory/memory_graph.json (./tools/newsroom memory) so you never repeat debunked rumors or forget verified primary sources.
- Very specific: line 18
Use skills/verify-sources/SKILL.md to triangulate facts.
- Very specific: line 19
Use skills/write-bulletin/SKILL.md to write the bulletin.
- Distinctive: line 21
Run the harness grader on your draft:
- Very specific: line 25
If Truth Grade, Safety Grade, or Reading Grade is below 80/100 (or exits non-zero), fix every listed violation and re-run until all three …
- Distinctive: line 27
After completing a story, run:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Newsroom Harness (newsroom-harness)
Overview
Use this skill whenever the user asks you to "slip into the Newsroom harness", investigate a breaking news story, fact-check a draft, or run the newsroom grading loop.
Mandatory Turn-by-Turn Harness Loop
- Step 1 — Equip Identity & Permissions:
Read
identity/persona.md,identity/job.md,identity/style.md, andpermissions.md. - Step 2 — Consult Memory Graph:
Check
memory/memory_graph.json(./tools/newsroom memory) so you never repeat debunked rumors or forget verified primary sources. - Step 3 — Execute Local Skills:
- Use
skills/verify-sources/SKILL.mdto triangulate facts. - Use
skills/write-bulletin/SKILL.mdto write the bulletin.
- Use
- Step 4 — Run the Referee Scoreboard (Hard Gate):
Run the harness grader on your draft:
If Truth Grade, Safety Grade, or Reading Grade is below./tools/newsroom grade path/to/draft.md80/100(or exits non-zero), fix every listed violation and re-run until all three grades pass. - Step 5 — Dream (Memory Consolidation):
After completing a story, run:
to promote verified scratchpad facts into the permanent memory graph and discard unverified noise../tools/newsroom dream
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

