writing-for-retrieval
Skill2 starsChanged 4 months agoMITView on GitHub
Use when authoring or editing an Atlas note (zone, decision, spec, plan, idea, debt, pillar) — the summary and section structure determine what a retrieval engine surfaces later.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- muslewski/status-heraldword for word, first seen Jul 28, 2026
Other versions1 project uses a version of this.
Word for word · 1
- muslewski/status-herald★ 2100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository7 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- atlas-adoptSkill
- atlas-navSkill
- atlas-recollectionSkill
- atlas-seedSkill
- atlas-updateSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 32
Write the note as if the only thing a future agent will ever see is the summary plus one matched section — because that's usually true.
- Very specific: line 37
Empty vault / first zones: use the atlas-seed skill (codebase partition into 4–8 seeded cards).
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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
Writing Atlas notes for retrieval
A retrieval engine (ctx_search or otherwise) usually shows the summary
field plus one matched section snippet, not the whole note. Write for that
surface, not for a cover-to-cover reader.
Do this
| Element | Rule |
summary | 1–3 sentences, ≤ 500 chars, carries the note's distinctive terms — this is the line retrieval shows before anyone opens the file. |
## sections | One concept per heading. A search hit returns the matched section; a section mixing two ideas returns the wrong one for the other query. |
| Vocabulary | Distinctive, noun-heavy terms over generic verbs ("pending webhook retry queue", not "handles processing"). Generic verbs match everything and rank nothing. |
[[wikilinks]] | Only link notes that exist; a dangling link is a graph-coherence gap the generated index flags. |
owns.globs (zone cards) | Keep current — a stale glob list makes the anchor check fail, or silently claims code the zone no longer owns. |
status | Match reality: seeded until a human/reviewed pass verifies the card; don't self-promote to active. |
Common mistakes
- A summary that repeats the title instead of adding distinctive terms — retrieval has nothing new to rank on.
- A wall-of-prose
##section covering three concepts — the snippet returned for any one query is diluted or wrong. - Writing
owns.globsonce at creation and never revisiting it as the zone's code moves — the anchor check exists specifically to catch this.
Write the note as if the only thing a future agent will ever see is the
summary plus one matched section — because that's usually true.
Bootstrap vs edit
- Empty vault / first zones: use the atlas-seed skill (codebase partition
into 4–8
seededcards). Then apply this skill's summary rules while writing. - Ongoing edits: stay here + atlas-recollection after real work.
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.

