AGENTS.md vs CLAUDE.md vs GEMINI.md
Every comparison page on the web is a capability table someone wrote from the vendor docs in an afternoon. None of them can tell you how many real repositories ship each file, how long the median one runs, or how often two of them turn up in the same project. Those answers need a corpus, and RuleStack counts one every night.
| Format | AGENTS.md 2026-08-02 | CLAUDE.md 2026-08-02 | GEMINI.md 2026-08-01 |
|---|---|---|---|
| Asserted — what the docs say it can express | |||
| File | AGENTS.md | CLAUDE.md | GEMINI.md |
| Status | open-standard | vendor | vendor |
| Cross-tool | Yes | No | No |
| Tools that read it |
|
|
|
| Frontmatter | No | No | No |
| Glob targeting | No | No | No |
| Imports other files | No | Yes | Yes |
| Nested / per-directory | Yes | Yes | Yes |
| Multiple files | Yes | Yes | Yes |
| User-level scope | No | Yes | Yes |
| When it applies | Always in scope; in a monorepo the nearest AGENTS.md to the edited file wins. | Always in scope. Project file, plus nested files in subdirectories, plus a user-level file (~/.claude/CLAUDE.md) that applies across every project. | Hierarchical memory: a global ~/.gemini/GEMINI.md, then the project file, then nested files, concatenated from broadest to narrowest. |
| Size limits | None | None | None |
| Measured — counted from real repositories on 2026-08-04 | |||
| Repositories carrying it | 399 | 248 | 33 |
| Share of indexed repos | 57.2% | 35.5% | 4.7% |
| Files indexed | 1,373 | 775 | 46 |
| Median length (words) | 953 | 2,042 | 257 |
| 90th percentile length | 2,956 | 2,998 | 1,704 |
| Files naming commands | 42.8% | 34.2% | 52.2% |
| Files containing code | 36.0% | 34.5% | 45.6% |
| Median quality score | 52 | 44 | 60 |
| Median repository stars | 9,402 | 114 | 11,320 |
How often they turn up together
A capability table cannot answer this at all. Of the repositories carrying the first format, this is the share that also carry the second — counted, both directions, on 2026-08-04.
CLAUDE.md + GEMINI.md
2026-08-0417
repositories carry both. That is 6.9% of everything with CLAUDE.md, and 51.5% of everything with GEMINI.md.
AGENTS.md + GEMINI.md
2026-08-0423
repositories carry both. That is 5.8% of everything with AGENTS.md, and 69.7% of everything with GEMINI.md.
AGENTS.md + CLAUDE.md
2026-08-04126
repositories carry both. That is 31.6% of everything with AGENTS.md, and 50.8% of everything with CLAUDE.md.
What each one is for
AGENTS.md
2026-08-02A plain-markdown README for coding agents, deliberately unopinionated: no frontmatter, no globs, no vendor keys. That minimalism is why it became the one file a dozen different agents will read, and why it carries the least per-file targeting power of any format here.
Strengths
- Read by more agents than any other format — one file covers most of a team's tooling
- No schema to get wrong; a plain markdown file always parses
- Nested files let a monorepo scope instructions per package
Weaknesses
- No glob targeting — every rule is loaded for every file, spending context
- No imports, so shared rules must be duplicated across packages
- No frontmatter means no machine-readable metadata for tooling to act on
CLAUDE.md
2026-08-02Claude Code's memory file. Shaped like AGENTS.md but with two things it lacks: @path imports, so shared rules live in one place, and a user-scope layer that follows the developer across repos rather than shipping with the code.
Strengths
- @path imports — shared conventions written once and pulled into each file
- A user-scope file lets personal preferences live outside the team repo
- Nested project files compose with the root file rather than replacing it
Weaknesses
- Single-vendor: no other agent reads it, so multi-tool teams maintain two files
- No glob targeting, same context cost as AGENTS.md — path-scoped rules exist, but in a separate .claude/rules/ file, not in CLAUDE.md itself
GEMINI.md
2026-08-01Gemini CLI's memory file, structurally close to CLAUDE.md — @imports and a user-scope layer — which is why repos that carry both usually carry near-identical text in each.
Strengths
- @path imports and a global user file, same composition model as CLAUDE.md
- Explicit hierarchy from global to local
Weaknesses
- Google-only
- No glob targeting
