lktieplktiep/cortex-hub/.vscode/copilot-instructions.md

cortex-hub

Copilot instructions105 starsChanged 34 days agoMITView on GitHub

At the START of every conversation: 1. Call cortexsessionstart with repo: "https://github.com/lktiep/cortex-hub.git", agentId: "vscode-copilot", mode: "development" 2. cortexknowledgesearch + cortexmemorysearch once — recall what the last session decided Search once, read all ten. On the cortex-hub index the target file is in the top 10 for 15/15 queries but at rank 1 for only 8/15 — so scan the whole result set, and never re-run a reworded version of the same query. Ask a different question or switch tool instead.

We split this copilot-instructions.md into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 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.
3
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
Call cortex_session_start with repo: "https://github.com/lktiep/cortex-hub.git", agentId: "vscode-copilot", mode: "development"
line 5
Very specific:
Run verify commands from .cortex/project-profile.json, then cortex_quality_report.
line 34
Very specific:
After a push: cortex_code_reindex.
line 35

≈ “Common” is an estimate: at least 1 in 5 of the 3,669 Copilot instructions 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.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

copilot-instructions.md

Cortex Hub — Auto MCP (MANDATORY)

At the START of every conversation:

Call cortex_session_start with repo: "https://github.com/lktiep/cortex-hub.git", agentId: "vscode-copilot", mode: "development"
cortex_knowledge_search + cortex_memory_search once — recall what the last session decided

Finding code — start from what you know, not from a fixed ladder

You already know	Start with
A symbol name	`cortex_code_context(name)` — exact graph lookup, plus callers/callees/imports in one call
Only the behaviour	`cortex_code_search(query, limit: 10)` — ranked hybrid search, one call
An exact literal (env var, config key, error string)	`rg` / `grep` — this is not a ranking problem
A relationship across files	`cortex_cypher`

Search once, read all ten. On the cortex-hub index the target file is in the top 10 for 15/15 queries but at rank 1 for only 8/15 — so scan the whole result set, and never re-run a reworded version of the same query. Ask a different question or switch tool instead.

cortex_code_impact before editing something exported or shared. Knowledge and memory are for errors and decisions, not for locating code.

Across repos: omit repo: to search every repo in the organization of this project (the client + server + tools of one product). Projects are isolated per organization, so it never reaches another one. Running sessions in two organizations with one API key? Pass org:.

Error Protocol

cortex_knowledge_search first — someone may have solved this
cortex_memory_search — you may have seen it before
Fix the error
Non-obvious fixes: cortex_knowledge_store

Quality Gates

Run verify commands from .cortex/project-profile.json, then cortex_quality_report. After a push: cortex_code_reindex. End session: cortex_session_end with sessionId and summary.

Download now (ZIP)

Where files are headingNew Copilot instructions files have more rules about branches and pull requests, and fewer about code style.
Of the Copilot instructions files we read, the 2,297 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository2 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

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.

Check your setup, free