How do I install the Pydantic AI Foundation skill?

Run npx skillmds@latest add ryan-brosas/pydantic-ai-foundation in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Pydantic AI Foundation safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Pydantic AI Foundation?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Pydantic AI Foundation free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Pydantic AI Foundation?

ryan-brosas (@ryan-brosas) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Pydantic AI Foundation by ryan-brosas

Name: Pydantic AI Foundation
Author: ryan-brosas

Pydantic AI Foundation

Use when porting or building an agent runtime's core machinery — deferred/external tool calls, human-in-the-loop approval, pause/resume envelopes, barrier-segmented parallel execution, structured-output schema resolution, capability middleware with ordering constraints, toolset composition (wrapper/combined/prefixed/renamed/prepared/filtered/dynamic/approval/external), streaming partial→final validation, the durable-execution shared kernel (string-only model round-trip with credential-leak rejection, replay-unit enqueue/cancel guards, live-vs-replay event-stream split), the durability-engine adapter plane (anyio-shielded cancel forwarding, boundary-guarded context rehydration, sequence-keyed cache keys, legacy-replay dual paths), static tool-choice deadlock gates, transport-level tenacity retries honoring Retry-After, session cancellation as typed resumable exceptions, and replayed-history hardening for partially-rejected provider turns. Source code and direct tests are ground truth.

ryan-brosas Updated Sep 17, 2026

File contents

pydantic-ai: Agent Runtime Foundation

Foundation contract

This is a cold, source-specific foundation, not an operational procedure. It is historical, revision-pinned evidence rather than timeless guidance. Current project source, tests, requirements, and runtime behavior outrank this projection and decide what ships.

Provenance and revision

Source record: pydantic-ai; portable upstream identity, license, coverage caveats, and retrieval history are preserved in the index.
Recorded revision: main@a5b5fb7a247f863599d61dfa9159bc2ebc786255.
Full provenance, coverage caveats, boundaries, and preserved evidence: references/index.md.

Topic map

Representative topics: Deferred envelope; Step-final deferral ladder; Approval dispatch; End strategies; Parallel execution; Retry budget; Availability refusals; Declarative deferral.
Scope and retrieval questions: references/index.md#use-this-for.
Complete capsule chooser: references/index.md#load-the-matching-source-dump.
Detailed grouped map: references/index.md#capsule-map.
Source limitations and portability boundaries: references/index.md#boundaries.

Retrieval

Open the index, choose one capsule matching the active question, and load only that capsule. Revalidate its cited source and revision before relying on it. Do not bulk-load the inventory, treat historical claims as current truth, or apply this foundation as a procedure.

---name: pydantic-ai-foundationdescription: Use when porting or building an agent runtime's core machinery — deferred/external tool calls, human-in-the-loop approval, pause/resume envelopes, barrier-segmented parallel execution, structured-output schema resolution, capability middleware with ordering constraints, toolset composition (wrapper/combined/prefixed/renamed/prepared/filtered/dynamic/approval/external), streaming partial→final validation, the durable-execution shared kernel (string-only model round-trip with credential-leak rejection, replay-unit enqueue/cancel guards, live-vs-replay event-stream split), the durability-engine adapter plane (anyio-shielded cancel forwarding, boundary-guarded context rehydration, sequence-keyed cache keys, legacy-replay dual paths), static tool-choice deadlock gates, transport-level tenacity retries honoring Retry-After, session cancellation as typed resumable exceptions, and replayed-history hardening for partially-rejected provider turns. Source code and direct tests are ground truth.---# pydantic-ai: Agent Runtime Foundation## Foundation contractThis is a cold, source-specific foundation, not an operational procedure. It ishistorical, revision-pinned evidence rather than timeless guidance. Currentproject source, tests, requirements, and runtime behavior outrank thisprojection and decide what ships.## Provenance and revision- Source record: `pydantic-ai`; portable upstream identity, license, coverage caveats, and retrieval history are preserved in the index.- Recorded revision: `main@a5b5fb7a247f863599d61dfa9159bc2ebc786255`.- Full provenance, coverage caveats, boundaries, and preserved evidence: `references/index.md`.## Topic map- Representative topics: Deferred envelope; Step-final deferral ladder; Approval dispatch; End strategies; Parallel execution; Retry budget; Availability refusals; Declarative deferral.- Scope and retrieval questions: `references/index.md#use-this-for`.- Complete capsule chooser: `references/index.md#load-the-matching-source-dump`.- Detailed grouped map: `references/index.md#capsule-map`.- Source limitations and portability boundaries: `references/index.md#boundaries`.## RetrievalOpen the index, choose one capsule matching the active question, and load onlythat capsule. Revalidate its cited source and revision before relying on it. Donot bulk-load the inventory, treat historical claims as current truth, or applythis foundation as a procedure.

Frequently asked questions