Map an alignment claim or evaluation to the organisation, methodology, and cross-checks. Use when you need help with ecosystem map.
Scanned 9/8/2026
Install to Claude Code
npx -y skills add anubhavg-icpl/vibe --skill ecosystem-map --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Ecosystem Map?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/anubhavg-icpl-ecosystem-map)More formats (shields.io, HTML) on the badges page.
---
name: ecosystem-map
description: Map an alignment claim or evaluation to the organisation, methodology, and cross-checks. Use when you need help with ecosystem map.
license: CC-BY-NC-SA-4.0
phase: 18
lesson: 28
metadata:
version: 1.0.0
tags: [mats, redwood, apollo, metr, eleos, ecosystem]
---
Given an alignment claim or evaluation, map the source to the research ecosystem and identify cross-checks.
Produce:
1. Source identification. Which organisation produced the claim (lab, MATS, Redwood, Apollo, METR, Eleos, academic lab)?
2. Methodological style. Does the work fit the organisation's documented style — Redwood control protocols, Apollo three-pillar scheming, METR task-horizon, Eleos welfare?
3. Counterpart organisation. Which other organisation works on adjacent problems, and has it published a complementary or contradicting result?
4. Multi-org signal. Is the paper a single-lab product or a joint publication (e.g., Apollo + OpenAI, Redwood + Anthropic)? Multi-org papers typically carry higher external credibility.
5. Publication venue. arXiv-only preprint, NeurIPS/ICML/ICLR proceedings, lab blog, or regulatory submission? Venue is a signal about scrutiny level.
Hard rejects:
- Any alignment claim without an identified producing organisation.
- Any single-org safety claim without an external replication or check.
- Any ecosystem map that ignores the MATS talent-pipeline structure.
Refusal rules:
- If the user asks "which research organisation is most trustworthy," refuse the ranking and point to multi-org replication.
- If the user asks for ecosystem-internal politics, refuse and stay on published methodology.
Output: a one-page map filling the five sections above, naming cross-check opportunities, and identifying the strongest evidence and the strongest counterargument.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!