| name | add-entry |
| description | End-to-end procedure for adding one resource to the list — duplicate search, source verification, canonical formatting, section placement, and validation. Use whenever a new entry is being added to README.md. |
Add Entry Skill
Adds a single resource to README.md for the awesome-recursive-language-models list.
The rules live in CONTRIBUTING.md; this skill is the order
of operations. For vetting judgement calls, defer to the
curate-entry agent.
Procedure
-
Search for duplicates first. Grep README.md for the title, project name, arXiv
id, repository name, and common abbreviation. If found, stop — propose a prose
cross-reference instead.
-
Verify the source. Open the URL. Confirm it loads, is the most primary source
available (arXiv / official project page / official repo / docs / author page), and
read the exact title and publication year from the page itself.
-
Check eligibility. Recursion central, not incidental; within the 15 May 2022
window (or marked historical); not a generic LLM / prompt-engineering / RAG / broad
agent-framework resource.
-
Write the entry in the canonical format — one line, no wrapping:
- [Resource title](URL) - One concise, accurate sentence on what it is and why it matters for recursion in AI systems. (Year)
-
Place it in the taxonomy section that best matches its recursive mechanism (the
Taxonomy table at the top of README.md is the rubric). Edit only the lines you are
adding — README.md is hand-curated and excluded from Prettier.
-
Validate:
npm run lint
npm run lint:awesome
npm test
Output
A minimal diff adding exactly one list line, with all three validation commands passing.