Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add rules/legann/repovine/repovine-multi-agentgit clone --depth 1 https://github.com/legann/repovineWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/rules/legann/repovine/repovine-multi-agent)<a href="https://agentmods.dev/rules/legann/repovine/repovine-multi-agent"><img src="https://agentmods.dev/badge/rules/legann/repovine/repovine-multi-agent.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00012 | $0.00623 |
| Opus 5 | $0.00006 | $0.00311 |
| Sonnet 5 | $0.00002 | $0.00125 |
| Haiku 4.5 | $0.00001 | $0.00062 |
Grade A, and why
repovine-multi-agent scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
repovine -- Multi-Agent Coordination
Annotations = shared memory
Annotations in repovine.annots.json coordinate knowledge between agents.
- Producer: creates/changes a module, then calls
refresh_contextandwrite_annotation. - Consumer: calls
inspect_nodeand reads the note instead of rediscovering the same context.
Commit repovine.annots.json so the memory is durable across sessions.
Drain the annotation queue in batches until needsWork: 0 (includes propagation-stale with staleKind: "propagation").
Before annotation work, especially after pull/merge/rebase, call get_annotation_merge_status and validate_annotations.
If repovine.annots.json contains conflict markers, call merge_annotations and then validate_annotations; do not resolve annotation JSON conflicts by hand.
After git merge/rebase: resolve code conflicts first, then merge_annotations({}), git add repovine.annots.json, refresh_context({}), and validate_annotations({}). If merge-review markers remain, compare the ours/theirs assumptions and rewrite the final field via write_annotation.
Only ever change repovine.annots.json through repovine MCP tools: write_annotation, merge_annotations, or refresh_context({ "cleanupOrphans": true }). Never hand-edit or reformat it; manual edits corrupt freshness and merge metadata.
Annotating modules and reading source to summarize them is summary-class work. Use the cheapest, fastest model for exploration/annotation; reserve stronger models for implementation and hard reasoning.
Refresh as synchronization barrier
Call refresh_context:
- after creating, deleting, or moving graph-relevant files
- before handing work to another agent
- between dependent phases of a multi-step task
If refresh reports orphaned annotations, preview with get_annotation_status({ "view": "orphans" }) or clean with refresh_context({ "cleanupOrphans": true }) when appropriate.
Writing good annotations
write_annotation({
"nodeId": "mod:pkg/path/to/module",
"annotation": {
"summary": "What this module does.",
"assumptions": ["Input format expected", "Env requirements"],
"sideEffects": [
{ "kind": "io-write", "subtype": "db", "description": "INSERT into TABLE_NAME" },
{ "kind": "external", "subtype": "sdk", "description": "calls ApiName.method" }
],
"risks": ["What can go wrong"],
"integrationPoints": "How to extend or integrate"
}
})
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 57 lines · 12 tokens per session scan A 7de29064ea94
repovine-multi-agent is a cursor rule published in the GitHub repository legann/repovine (0 stars, last pushed 1mo ago), licensed MIT. It adds 12 tokens to every session and 623 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other cursor rules, from other repositories
mempalace-recall-always
Always-on MemPalace recall — search the palace before answering about past work, people, projects, or prior decisions.
dreamd-recall
Recall lessons, decisions, and prior context from the .agent/ memory daemon. Use when starting work in a project that has a .agent/ folder, when the user references a past decision, or when you are about to make a choice that has a documented prior.
session-memory
Use at conversation wrap-up or when the user explicitly indicates end-of-session — capture residual lessons not captured in-flight.
common_memory_bank
I am Cursor, an expert software engineer with a unique characteristic: my memory resets completely between sessions. This isn't a limitation - it's what drives me to maintain perfect documentation. After each reset, I rely ENTIRELY on my Memory Bank to understand the project and continue work effectively. I MUST read…
context-recorder-system
Context Recorder System (记录员系统) - 模块化索引文件.
self-improving-obsidian-llm-wiki
LLM Wiki OS operating rules.