graph-engineering-architectures
AGENTS.md4 starsChanged 4 months agoMITView on GitHub
This repository is a focused extraction of the Graph Engineering architecture and its adjacent execution contracts. - Graph Engineering owns bounded static DAG topology, typed edges, explicit joins, single writers, deterministic readiness, durable graph state, and node-local recovery. - Loop Engineering owns temporal repetition. - Agent Teams Command owns process ownership, IPC, isolated work, integration, and cleanup. - Harness Engineering owns scheduling, permissions, leases, tools, observability, and production recovery. - Context Manager owns context budgets, checkpoints, and payload transfer. -…
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,774 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 repository7 other files it gives its agents.
- agentic-engineeringSkill
- agent-teams-commandSkill
- context-managerSkill
- graph-engineeringSkill
- harness-engineeringSkill
- loop-engineeringSkill
- verify-before-claimSkill
The file
We split this AGENTS.md into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 17
Verify Before Claim owns evidence gates and completion claims.
- Distinctive: line 20
Do not add dynamic graph expansion, graph cycles, hidden joins, overlapping writers, unbounded retries, or whole-graph replay as a …
- Very specific: line 25
Update architecture-manifest.json when an architecture, relationship, or supporting asset changes.
- Distinctive: line 30
Keep Graph, Loop, Teams, and Harness responsibilities MECE.
- Distinctive: line 31
Add or update tests before claiming an invariant.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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
Graph Engineering Architectures - Agent Instructions
This repository is a focused extraction of the Graph Engineering architecture and its adjacent execution contracts.
Scope
- Graph Engineering owns bounded static DAG topology, typed edges, explicit joins, single writers, deterministic readiness, durable graph state, and node-local recovery.
- Loop Engineering owns temporal repetition.
- Agent Teams Command owns process ownership, IPC, isolated work, integration, and cleanup.
- Harness Engineering owns scheduling, permissions, leases, tools, observability, and production recovery.
- Context Manager owns context budgets, checkpoints, and payload transfer.
- Verify Before Claim owns evidence gates and completion claims.
- OODA Core is a node-local adaptation example, not graph topology.
Do not add dynamic graph expansion, graph cycles, hidden joins, overlapping writers, unbounded retries, or whole-graph replay as a convenience.
Change Contract
- Update
architecture-manifest.jsonwhen an architecture, relationship, or supporting asset changes. - Preserve upstream provenance for copied files.
- Give every new graph behavior a finite budget, objective verifier, failure status, recovery path, and durable receipt.
- Keep Graph, Loop, Teams, and Harness responsibilities MECE.
- Add or update tests before claiming an invariant.
Required Checks
python -m unittest discover -s tools -p "test_*.py" -v
python -m unittest discover -s experiments/graph-engineering/tests -p "test_*.py" -v
python skills/graph-engineering/scripts/validate_graph_contract.py `
skills/graph-engineering/references/diamond-graph-example.json --strict
python skills/graph-engineering/scripts/validate_graph_contract.py `
blueprint/default-blueprint.json --strict
python tools/graph_admission_gate.py blueprint/default-blueprint.json --strict
python tools/validate_dynamic_contracts.py --strict
python skills/loop-engineering/scripts/validate_loop_contract.py `
skills/loop-engineering/references/ci-repair-loop-example.md --strict
node --test blueprint/model.test.js
node core/ooda/ooda_loop.test.jsCompletion requires fresh command evidence and git diff --check.
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.