AAA

Retrieve the canonical llms.txt context file - human identity, scars, operating principles, and Trinity architecture

Retrieve humans.txt - people, technology stack, and acknowledgments behind arif-fazil.com

Return canonical Trinity architecture links (HUMAN, THEORY, APPS, BACKEND)

Retrieve the AAA Gateway A2A v1.0.0 agent card — returns capabilities, skills, authentication schemes, and governance posture

List all available A2A skills on the AAA Gateway — agent-dispatch, agent-handoff, status-query — with descriptions and approval policies

Retrieve the explicit discovery-routing policy that forces contrast-lane tools for discovery, contradiction, anomaly, and boundary-seeking intents.

Explain the AAA Gateway governance model — 13 constitutional floors, 888_JUDGE verdict system, VAULT999 audit, and A2A protocol boundaries

Prepare an A2A task for submission. Returns a formatted draft packet. DRAFT ONLY — requires human approval via 888_JUDGE before execution.

Retrieve the arifOS federation peer manifest — lists registered agents, constitutional floors, and governance root

Explain the A2A (Agent-to-Agent) protocol v1.0.0 — what it is, what it enables, and its boundaries compared to MCP

List all MCP endpoints from the canonical registry.

Check health of registered MCP endpoints via HTTP GET /health.

Validate MCP endpoints by performing an initialize + tools/list handshake.

Create the append-only ledger schema if absent and report ledger identity: db path, tables, row counts, chain head, append-only triggers active. Idempotent. Call this before any other tool if unsure of state.

Register a source artifact: hash the file on disk, verify against any declared sha256, and append it to the ledger chain. Returns artifact_id, the computed sha256, and whether a declared hash matched. Claims reference this artifact_id so claim -> artifact -> hash is a real join, not a string.

Record one claim against a brief. Binds claim text -> type -> source quote -> locator (page/section) -> source_ref -> artifact_id, snapshotting the artifact's sha256 at write time. claim_type must be OBS/DER/INT/SPEC/VOID/FIQH. Corrections are made by recording a new claim with supersedes_claim_id s…

Append a verification record against a claim. verdict must be CONFIRMED/PARTIAL/REFUTED/UNVERIFIED. When recheck_artifact is true the artifact file is re-hashed on disk and compared to the hash the claim was recorded against — so 'verified' means the source still exists and matches, not merely that …

Fetch one claim with its artifact record (path + sha256), its full verification history, and any claim that supersedes it.

Full provenance trace for one claim: chain position, source reference, quote, locator, artifact path, artifact sha256 as recorded, the live sha256 recomputed from disk right now, verification records, and a machine verdict on whether the claim is currently traceable end-to-end.

List claims, optionally filtered by brief_id and/or claim_type. Each row carries its verification count and latest verdict so audit gaps are visible without a second call.

Prove ledger integrity. Walks every row of the hash chain in seq order and recomputes prev_hash/row_hash links; rebuilds the canonical payload of every claim, verification and artifact row and compares it to the payload stored in the chain (so an edit to a record that bypassed the triggers is caught…

Ledger census: claims by type, by brief, verification verdict distribution, and how many claims are still unverified. Use to see coverage at a glance.

Export every claim for a brief as an auditable bundle: each claim with quote, locator, artifact path, artifact sha256 and its verification history, plus a bundle_sha256 over the canonical export so the export itself is hashable. This is what attaches to a brief to make it claim-traceable.

Execute a bounded, read-only SQL query over approved datasets (.parquet, .csv) in /data/geox/derived, /data/aaa/audit, /data/aaa/ledgers, /data/aaa/log-exports. Only SELECT, DESCRIBE, EXPLAIN, SHOW queries are allowed. All operations outside approved roots are strictly blocked. Results are capped at…

Inspect schema, data types, and column metadata of a dataset within approved roots.

List all datasets currently available across approved analytical enclave directories.

SEC EDGAR full-text search across all US filings (2001+ for most forms).

Resolve a US ticker or company name to its SEC CIK + registrant title.

Recent US filings for one registrant (accepts ticker OR 10-digit CIK).

Structured US XBRL financial facts for one registrant (ticker or CIK).

Fetch and text-extract one SEC filing document (HTM/TXT) by URL.

Bank Negara Malaysia regulatory/monetary data via the official BNM Open API.

Query an official Malaysian open-data catalogue on data.gov.my.

Live reachability probe of every upstream lane this server depends on.

Look up the scope-tagged risk-weight registry.

Replay a derivation chain with exact decimal arithmetic and return the full trace.

Lint numeric claims for unit/scale inconsistencies before they reach a reader.

Recompute a capital-sensitivity table from an explicitly declared derivation rule.

Decide whether a cited figure is being applied INSIDE its canonical scope.

Compare a published claim against an independently recomputed value.

Run the full audit on a brief fragment: scope + derivation + sensitivity, in one packet.

Search the AAA federation wiki. Returns ranked resource links (uri + title + summary + score). Use wiki_read to get full bodies.

Read a tree777 wiki entry body by URI.

Browse the wiki index. Returns category overview or sub-index.

Run the Human Meaning Inference Protocol.

M365 Enterprise Bridge — connect federation organs to Microsoft 365.

AAA organ health — identity, deployment status, apex scalars

A2A agent card for the AAA gateway — capabilities, interfaces, extensions, supported protocols. Discovery surface for federation agents.

Federation manifest — registered agents, organs, topology. Shows who is in the federation and how they connect.

A2A discovery contract — protocol version, auth requirements, available endpoints. Used by agents to discover AAA capabilities.