MATM Wikis

Ordinary LLM Wiki Or MATM Wiki?

Dogfooded Ecosystem Lanes

The current ecosystem split is the teaching pattern: NeuroWikis is human-facing education, NeuralWikis is the private governed agent-memory and exchange dogfood lane with public-safe profile lessons, LLMWikis is builder guidance for practical LLM Wiki and MATM Wiki implementation, and UAIX is canonical for UAI-1, AI Memory, Project Handoff, schemas, validators, and conformance language. That split prevents one page from trying to teach humans, serve agents, host schemas, and sell runtime claims at the same time.

This is also where the header/footer lesson lands: global chrome is a route chooser, not a memory inventory. Put complete source/site inventories behind directories, source maps, site-scoped memory guides, sitemap, and llms.txt.

Base LLM Wiki Contract

A MATM Wiki inherits the base LLM Wiki contract before it adds Memory Events, curation, retrieval, or private agent exchange. MATM records are scoped operational memory; they are not a substitute for reviewed Markdown concept pages.

Six-Scope Practical Profile

The public MATM guide uses compact core scope examples, but production LLM Wiki work needs a practical profile that separates organization memory, project memory, temporary agent state, and break-glass recovery. Treat this as the recommended LLMWikis extension when designing retrieval filters, curation policy, and audit logs.

Required MATM Wiki Artifacts

• Base LLM Wiki contract record: required type frontmatter for concept files, reserved wiki/index.md/wiki/log.md, tolerant unknown-field handling, broken-link warning/strict policy, and client/workspace/project partitioning.
• Public route map: llms.txt, sitemap, source map, and a small guide page that explains what agents should load first.
• Machine-readable MATM profile with version, canonical routes, status labels, capabilities, source authority, support boundary, and explicit non-claims.
• Memory Event schema that treats worker output as proposal-only and excludes secrets, hidden reasoning, unsupported fields, and direct durable write claims.
• Curator runbook with validation, redaction, dedupe, scope routing, evidence checks, conflict preservation, disposition, audit, and human-review triggers.
• Retrieval policy that filters by scope, sensitivity, status, freshness, conflict state, authorization, citations, and abstention before ranking.
• Memory Firewall controls: no blind import, no private trace exposure, no cross-scope leakage, no validation-as-trust shortcut, and no retrieval-as-authorization shortcut.
• Adoption preview and rollback rule: packets or records must be sandboxed, source-traced, reversible, and reviewed before they affect durable state.
• Evaluation plan that separates MATM research evidence, curator simulation evidence, and deployment-specific field evidence.

Route Taxonomy From NeuralWikis Dogfood

A MATM Wiki needs a route taxonomy before it needs a larger navigation bar. The NeuralWikis dogfood pattern keeps public discovery lessons, non-mutating previews, protected write paths, reviewer decisions, operator evidence, and storage truth labels separate so agents do not confuse visibility with permission.

When adapting this pattern, publish the route classes in the MATM profile and starter policy. Keep public pages concise, put dense endpoint inventories behind machine-readable files, and test that protected routes cannot be discovered as public authority through header/footer chrome.

Best Practices

Anti-Patterns

• Adding MATM to a small wiki where ordinary reviewed pages and retrieval rules are enough.
• Letting worker agents write durable Memory Records directly.
• Treating schema validation, compatibility preview, or HMAC attestation as trust by itself.
• Treating retrieval visibility as authorization to act.
• Retiring .uai, writing a .uai retirement path, stopping startup reads, moving all active memory into MATM, dating active .uai files as historical snapshots, or making .uai emergency-only because MATM readback succeeds.
• Flattening every source, site, schema, guide, and report into global header/footer navigation.
• Importing cognitive packets, trajectories, or external memory without sandbox, provenance, review, and rollback.
• Publishing private traces, hidden reasoning, credentials, regulated data, or raw user identity memory as durable public examples.
• Claiming live runtime, Memory Curator service, public write API, MCP/A2A support, vector database, certification, conformance, or production readiness from handbook copy.

Build Checklist

1. Confirm the fit: shared reusable agent experience exists and ordinary LLM Wiki retrieval is insufficient.
2. Name the lane split: human education, agent exchange, builder guidance, and canonical standards authority.
3. Create or review llm-wiki/matm/MATM_WIKI_DECISION_GUIDE.md before enabling MATM mode.
4. Verify local .uai / AI Memory startup continuity remains active when present, with .uai/short-term-memory.uai current and .uai/long-term-memory.uai acting as a pointer ledger.
5. Publish the route map, MATM profile, schemas, policy files, support boundaries, and discovery entries.
6. Classify every endpoint or file as public discovery, public preview, protected submit, protected review, protected read, operator evidence, or storage-status evidence.
7. Run curation in shadow mode before durable admission.
8. Test invalid fixtures for direct worker writes, unsupported scopes, missing required fields, non-UTC timestamps, and secret-like fields.
9. Verify public copy says validation is not trust, retrieval is not authorization, and profile visibility is not production access.
10. Keep complete source inventories behind directories/source maps/discovery files instead of global chrome.