xmtp-signet
xmtp-signet contiene 7 skills recopiladas de galligan, con cobertura ocupacional por repositorio y páginas de detalle dentro del sitio.
Skills en este repositorio
Use the xmtp-signet to connect an agent to XMTP conversations through the v1 credential model. Covers the signet runtime, the owner/admin/operator/credential/seal hierarchy, the `xs` CLI surface for daily operation (status, chat, msg, inbox, search, lookup, consent, seal inspection, credential inspection, policy inspection), and how harnesses connect over WebSocket or MCP without direct XMTP access. Use this skill whenever someone asks what the signet is, how to run an agent on top of it, how to send or read messages through it, how to join or host Convos conversations, how to inspect seals or credentials, how reveals work, how trust is surfaced, or how a harness should stream and respond to XMTP traffic through the signet. For orchestrator-side setup — bootstrapping the daemon, creating operators, issuing credentials, managing keys and wallets — use the `xmtp-admin` skill.
Administer an xmtp-signet as an orchestrating agent or owner — bootstrap the daemon, initialize and rotate keys, create managed wallets, create and remove operators, define policies, issue and revoke credentials, create and link managed inboxes, verify seals and walk seal history, and drive the privileged `xs` command surface that sets agents up to operate against the signet. Use this skill whenever someone asks how to set up a signet for agents, how to issue or revoke a credential, how to provision an operator with a scoped policy, how to rotate or export operator keys, how to link an inbox to an operator, how to verify a seal chain, or how to run privileged admin flows. For day-to-day agent use (messaging, reading, inspecting credentials and seals, harness connection), use the `xmtp` skill.
Understand and use the xmtp-signet to connect agents to XMTP conversations through the current v1 credential model. Covers the signet runtime, the owner/admin/operator/credential/seal hierarchy, permission scopes, reveals, and how harnesses connect through WebSocket or MCP without direct XMTP access. Use this skill whenever someone asks what the signet is, how to scope an agent's permissions, how to connect an agent through the signet, what seals communicate, how trust and verification work, how to deploy a signet, or how an agent should participate in XMTP conversations through the signet.
Work on the xmtp-signet codebase — add features, fix bugs, write handlers, extend transports, create schemas, and understand the current v1 architecture. Teaches the handler contract, package tiers, error taxonomy, Result types, and testing patterns. Use this skill whenever working on any packages/* code, adding a new feature to the signet, writing or modifying a handler, creating or updating Zod schemas, extending a transport adapter, debugging signet internals, understanding how the packages relate, or asking "where does this code go?"
Understand and navigate the libxmtp Rust codebase — the core XMTP protocol library that implements MLS-based messaging. Use this skill when analyzing libxmtp's architecture, understanding how it maps to the JS/mobile/WASM SDKs, identifying modularization opportunities, or tracing how a specific feature flows through the crate structure. Reference material lives in .reference/libxmtp/ (shallow clone).
Run end-to-end tracer bullets against the real xmtp-signet codebase. Executes user stories step-by-step, pausing at failures to diagnose and fix. Use when testing a user story, running an end-to-end flow, validating that things actually work, doing a tracer bullet, smoke testing, or when the user says 'let's try it' or 'does this actually work'. Proactively use this when a feature has been implemented but never run for real.
Look up current XMTP documentation using the blz CLI. Use when you need to check XMTP SDK methods, group permissions, identity model, content types, MLS details, or any XMTP protocol or API question. Prevents hallucinating outdated SDK patterns. Use when: (1) looking up XMTP SDK methods or patterns, (2) verifying API signatures before writing code, (3) understanding XMTP protocol concepts (groups, MLS, inboxes, attestations), (4) checking content type schemas, (5) any question about how XMTP works.