← all claims · /methodology#for-ai-agents

xrpl.methodology_for_ai_agents.methodology_mcp_envelope_anchors_resolve

Every methodology_url cited in an MCP tool envelope resolves to a real element on /methodology (contract per AGENT_TIER_DESIGN.md:410).

yellow Public-infrastructure dependent OR no independent cross-check path yet. Free-tier only until Batch B walker migration lands (post-soak, ~2026-08-31) and/or an independent verification walker is wired for this series.
Machine-readable status: /claims/xrpl.methodology_for_ai_agents.methodology_mcp_envelope_anchors_resolve.json or send Accept: application/json on this URL. URI is permanent — safe to cite.

What this claim says

label
Every methodology_url cited in an MCP tool envelope resolves to a real element on /methodology (contract per AGENT_TIER_DESIGN.md:410).
page
/methodology#for-ai-agents
claim id (yaml)
methodology_mcp_envelope_anchors_resolve
behavior
MANUAL_CURATION
Layer 3 source
self-attested-only

Risk note

Envelope contract from AGENT_TIER_DESIGN.md:410 ("Every methodology_url resolves. Simple HTTP head-check on all URLs referenced in envelopes."). Discovered violated during the 2026-07-31 anchor audit — 5 anchors were being emitted by Day 2/3 tools with no matching id on /methodology. Fixed by landing the "MCP tool envelope sources" h2 with 5 h3 anchor targets, extended to 9 with Day 4 (amm, token-attestation, rwa, mpt), and to 10 with Day 7 (signed-snapshot, singular — the sibling human-facing #signed-snapshots h2 further up the page carries the deep-dive on the chain design). If a future tool ships with a new methodology_url, the corresponding h3 must land in the SAME commit or the envelope's contract silently breaks. See sibling methodology_for_ai_agents_discovery_links_resolve.

Evidence — code paths that back this claim

Verify

Sovereignty classification derives from the claim's Layer 3 source per /methodology. Last verified 2026-08-05. This surface is documented in docs/PAID_MACHINE_TIER_DESIGN.md § 3.1.