AI & Agents MCP servers

Model-agnostic MCP control plane for HAI: bounded missions, owner-gated actions, drift checks.

Recovers intent from imperfect AI output: JSON, tool calls, names, paths and Markdown edits.

MCP server exposing 11 honesty-SDK kits (grounding, trust, audit, corroboration) as agent tools.

Local Phocinae decision-engine MCP stdio server: fail-closed approval gate + classify/route/score.

Typed AI judgments for agents: choice, score, and yes/no probabilities instead of prose.

Checks an AI agent's 'done' against the files and URLs it claims. Missing = FALSE_DONE.

MCP server that provides AI agent pattern expertise for designing agent systems

Auditable memory for AI agents: provable deletion, tamper-evident audit, time-travel recall.

OSS to create Personas, Skills, Templates, Agents, and Memories to customize your AI experience.

Persistent, local-first memory for AI coding agents. Remember, recall and forget across sessions.

Let coding agents talk: search every agent session's history and message live sessions.

Local-first MCP server for LLM output verification — risk signals, injection and PII checks.

Let one coding-agent CLI consult another: Codex, Claude Code, OpenCode, Copilot or Antigravity.

Local gut check before risky agent actions: calibrated safe/intended probabilities, no API calls.

Shared knowledge base + task queue so Claude, ChatGPT and other agents can work together safely.

Long-term memory for agentic tasks. It tells you when it doesn't know.

Local, payload-free AI cost visibility for agents. Query spend by model, route, or tagged work.

Run Codex and Claude Code on your computer from an MCP client, and track or continue their runs.

Event bus for AI agents: watch_topic long-polls live events; manage topics, schemas, webhooks.

Checks a remote MCP server against the Claude directory rules and guides its publication.

Buy human image preference comparisons on the Pebbler app with locally controlled USDC payments.

Agents can't skip steps: runs gated stepfiles on the client's own model.

MCP-native CEP for durable agents: event-time patterns, semantic predicates, derived events.

Shared, persistent memory for AI agents, with semantic recall and live updates.

OpenAI-compatible multi-model inference via ComputeFlux (Polkadot testnet, TEE-verifiable). stdio.

Read-only MCP server for reusable AI prompts and assistant blueprints.

Records your Claude Code, Codex and Pi history before deletion. Cited-or-abstain recall.

Read-only MCP bridge for alethech — verifiable, Ed25519-signed agent memory over stdio.