life-lago
Skill2 starsChanged 7 months agoMITView on GitHub
Agent-facing interface to the Lago knowledge substrate. Provides Karpathy-style wiki operations (ingest, search, read, write, lint, wake-up, query) on top of Lago's event-sourced, content-addressed filesystem.
In this repository12 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- axumSkill
- cargo-fuzzSkill
- cargo-macheteSkill
- event-sourcing-architectSkill
- protobufSkill
- rust-async-patternsSkill
- rust-best-practicesSkill
- tokio-concurrencySkill
- tokio-patternsSkill
- checkSkill
- release-checkSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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
- Very specific: line 36
When the Lago daemon isn't running, all operations fall back to direct filesystem access on markdown files.
- Distinctive: line 45
All write operations emit canonical aios-protocol EventKind events:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
life-lago -- Lago Knowledge Substrate Skill
Universal agent brain stem built on Lago's event-sourced persistence layer.
Operations
| Command | Description | Lago Primitive | Event Emitted |
lago wiki init | Initialize a wiki session | Create session + BlobStore + KnowledgeIndex | SessionCreated |
lago wiki ingest <path> | Ingest source document | lago-knowledge ingest -> BlobStore.put -> index rebuild | ObservationAppended |
lago wiki search <query> | Hybrid search (BM25 + graph) | KnowledgeIndex.search_hybrid() | -- |
lago wiki read <slug> | Read entity page with provenance | KnowledgeIndex.get_note() | -- |
lago wiki write <slug> | Create/update wiki page | BlobStore.put -> manifest update -> index rebuild | MemoryProposed -> MemoryCommitted |
lago wiki lint | Check health (contradictions, orphans, gaps) | KnowledgeIndex.lint() | StateEstimated |
lago wiki wake-up | Assemble L0+L1 context (~900 tokens) | MemCube::assemble() + generate_index() | -- |
lago wiki log | Show recent knowledge operations | Journal.read() filtered to knowledge events | -- |
lago wiki query <q> | Search, synthesize, optionally file back | search + optional MemoryCommitted | MemoryCommitted (if filed) |
lago wiki diff | Show what changed since last session | diff_manifests() | -- |
Three-Layer Model
| Layer | Directory | Lago Backing | Mutability |
| Raw Sources | raw/ | BlobStore (SHA-256 + zstd, immutable) | Never modified |
| Wiki | wiki/ | KnowledgeIndex + MemCubes | LLM-maintained via write |
| Schema | CLAUDE.md + .control/ | Policy gates + setpoints | Co-evolved by user + LLM |
Graceful Degradation
When the Lago daemon isn't running, all operations fall back to direct filesystem access on markdown files. The skill detects Lago availability by checking LAGO_URL or localhost:3001.
Access Model
- Life/Arcan agents: Native Rust access via lago-knowledge crate
- Any other agent (Claude Code, Codex, Cursor): This skill provides the same capabilities through CLI commands + markdown output
Event Protocol
All write operations emit canonical aios-protocol EventKind events:
ObservationAppendedfor new knowledge ingestedMemoryProposed->MemoryCommittedfor wiki page writesPolicyEvaluatedfor Nous gate scoring decisionsStateEstimatedfor lint health assessments
Integration with Control Kernel
The knowledge substrate feeds the control loop:
- Belief state includes wiki context from wake-up + relevant entity pages
- Nous evaluators score knowledge operations (freshness, coherence, coverage, provenance)
- EGRI can optimize scoring thresholds, search weights, and promotion criteria
- Policy gates enforce knowledge governance (no entity without provenance, no promotion below Nous gate)
Configuration
Environment variables:
LAGO_URL-- Lago daemon URL (default: http://localhost:3001)LAGO_DATA_DIR-- Data directory for standalone modeLAGO_WIKI_DIR-- Wiki directory for filesystem fallback
Dependencies
Rust crates used:
lago-knowledge(BM25, graph, lint, index, ingest)lago-core(MemCubes, events, cognitive types)lago-store(BlobStore, manifest diff)lago-journal(event persistence)
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.