Commit Graph

Select branches

Hide Pull Requests

acp-support

add-plan-debate

agents-md-no-private-test-imports

approvals-code-mode

billea/readme-pydantic-stack-footer

bump-pydantic-ai-2.14

bump-readme-pai-floor-2

bump_monty

cache-bust-guard

capability/adaptive-reasoning

capability/approval

capability/background-tools

capability/code-mode-dynamic-catalog

capability/compaction

capability/filesystem-shell

capability/filesystem-shell-v2

capability/guardrails

capability/knows-current-time

capability/memory

capability/planning

capability/repo-context-injection

capability/secret-masking

capability/session-persistence

capability/skills

capability/stuck-loop-detection

capability/subagent

capability/system-reminders

capability/tool-budget

capability/tool-error-recovery

capability/tool-orphan-repair

capability/tool-output-management

capability/verification-loop

caps-docs-links

chore/bump-vcrpy

chore/untrack-settings-local

clamp-oversized-messages

claude/absurd-durability-capability

claude/budget-disclosure

claude/checkpoints

claude/codemode-os-function-GK8ew

claude/coding-harness

claude/compaction-structural

claude/cross-model-history-label

claude/docling-markdownify-conversion-2o4e67

claude/exciting-turing-TmpC5

claude/localstack-aws-emulation-0owHz

claude/loop-detection

claude/managed-tool-capability-r2vtg

claude/modal-sandbox-capability

claude/permission-policy

claude/planning-capability

claude/quirky-pasteur-DMxCC

claude/readme-codemode-extra-fix

claude/ripgrep-filesystem-support-nYYt5

claude/spill-query-surface

claude/staleness-tracker

claude/subagents-capability

claude/terminal-bench-agent

claude/writing-style-no-em-dashes

cm-temporal

code-mode-final-output

code-mode-monty

code-mode-tool-search

codemode-alias

codemode-example

codex/aica-attribution

codex/describe-code-mode-metadata-selection

codex/exempt-pydantic-ai-upload-age

codex/fix-322-protected-walkers

codex/fix-issue-360-search-files-docstring

codex/harness-ce-guides

codex/issue-346-clarify-code-mode-return

codex/pr-354-memory

codex/readme-completed-capabilities

codex/skills-capability

compat-test-pyright-scope

docs/primitives-architecture

docs/readme-refresh

docs/review-fixes-and-shell-allowlist

docs/unified-harness-docs

drop-pai-git-pin

drop-safe-for-ci

dynamic-workflow

exa-capability

exec_env_code_mode

execution-environments-v1

experimental/confined-self-authoring

feat/macroscope-capability

feat/sandbox-no-wall-clock

feature/logfire-managed-prompt

fix-380-clear-tool-results-typed-parts

fix-guardrails-readme-parity

fix-limit-warner-empty-request

fix-release-permissions

fix/codemode-prompt-tool-surface

fix/remove-codeowners

fix/update-readme-pr-links

guardrails

handle-deferred-in-code-mode

harness-webfetch

infra/repo-setup

launch/deprecation-shim

launch/public-infrastructure

launch/publishing

launch/publishing-v2

launch/publishing-v3

launch/publishing-v4

launch/publishing-v5

launch/publishing-v6

launch/publishing-v7

launch/remove-deprecation-shim

launch/rename-to-pydantic-ai-harness

launch/skills

macroscope-ignore

main

managed-agent-spec

managed-components

monty-beta-0.0.19

on-run-error-snapshot

oversized-tool-output

package-skeleton

plan/issue-42

pydai-latest

readme-version-bump

remove-defer-loading-fallback

remove-expeerimental

revert-258-capability/filesystem-shell-v2

send-tweet-on-release

shell-env-control

sql-mode

step-persistence

subagents-contain-errors

subagents-run-controls

verify-tool-defs-in-durable-tests

vp/zizmor-yamlfmt

worktree-agent-aa3ee95439db547e1

#129

#130

#131

#132

#133

#134

#135

#136

#137

#138

#139

#140

#153

#154

#155

#156

#157

#158

#159

#160

#161

#162

#163

#164

#165

#166

#167

#168

#168

#169

#169

#170

#170

#171

#171

#172

#172

#173

#173

#174

#174

#175

#175

#176

#177

#177

#178

#178

#179

#180

#180

#181

#181

#182

#183

#183

#184

#184

#185

#185

#186

#186

#187

#188

#189

#190

#191

#192

#194

#197

#198

#199

#200

#201

#202

#203

#204

#205

#206

#207

#208

#209

#210

#211

#212

#216

#217

#217

#219

#220

#222

#222

#223

#224

#225

#226

#228

#229

#231

#231

#234

#236

#237

#238

#240

#241

#242

#243

#244

#244

#245

#247

#249

#251

#256

#257

#258

#259

#260

#261

#262

#263

#264

#266

#267

#268

#268

#269

#269

#270

#271

#271

#272

#273

#274

#275

#280

#283

#284

#286

#287

#287

#288

#290

#291

#292

#293

#294

#295

#296

#297

#298

#299

#300

#303

#304

#305

#309

#311

#312

#316

#317

#319

#323

#323

#324

#325

#326

#327

#327

#328

#328

#329

#331

#331

#332

#332

#333

#333

#334

#334

#335

#335

#336

#336

#337

#337

#338

#338

#339

#339

#340

#340

#341

#342

#343

#343

#344

#344

#345

#345

#347

#348

#350

#350

#351

#352

#352

#353

#353

#354

#355

#355

#356

#356

#358

#359

#361

#362

#364

#364

#366

#366

#368

#370

#372

#372

#374

#374

#376

#377

#379

#379

#381

#381

#383

#383

#384

#386

#387

#388

#389

#390

#390

#391

#396

#396

#397

#398

#398

#399

#399

#87

#88

#96

#97

#98

pydantic-harness-v0.1.1

v0.0.1

v0.1.0

v0.1.1

v0.1.2

v0.2.0

v0.2.1

v0.3.0

v0.4.0

v0.5.0

v0.6.0

v0.7.0

v0.7.1

v0.8.0

Mono Color

• c19467beee Make the Modal session deal in bytes, decode in the toolset Aditya VardhanandClaude Opus 4.8 2026-06-24 17:25:26 +05:30
• 905761dd6d Resolve relative file paths against the sandbox working directory Aditya VardhanandClaude Opus 4.8 2026-06-24 17:20:59 +05:30
• f12b604ef8 Use Modal's native async filesystem API for ModalSandbox Aditya VardhanandClaude Opus 4.8 2026-06-24 16:55:25 +05:30
• bb6da9948d Keep framework tools native in CodeMode (#296) 3 people 2026-06-23 20:44:17 +08:00
• 6232e0762b feat(subagents): replace name-keyed mappings with a SubAgent list (#298) 4 people 2026-06-23 14:05:38 +02:00
• 50638595fd Add "Part of the Pydantic Stack" footer to README (#297) 3 people 2026-06-20 20:39:27 +01:00
• ae37c3e07b Add "Part of the Pydantic Stack" footer to README billea/readme-pydantic-stack-footer Bill EastonandClaude Opus 4.7 2026-06-20 14:17:02 -05:00
• 3f4cbd0a66 fix(code_mode): tighten run_code prompt against phantom imports and bad inputs (#295) 3 people 2026-06-19 14:06:03 -05:00
• 9486dbc957 feat(code_mode): append via instructions, replace via dangerously_replace_instructions fix/codemode-prompt-tool-surface David SFandClaude Opus 4.8 2026-06-19 13:49:47 -05:00
• 6e4158c621 refactor(code_mode): instructions is a plain str + export the default prose David SFandClaude Opus 4.8 2026-06-18 16:59:18 -05:00
• 3da572e785 feat(code_mode): make the run_code prompt replaceable via instructions David SFandClaude Opus 4.8 2026-06-18 12:09:39 -05:00
• 8cb2a5c5fa fix(code_mode): tighten run_code prompt against phantom imports and bad inputs David SFandClaude Opus 4.8 2026-06-18 11:14:29 -05:00
• 1b709d8e30 fix(overflow): read_tool_result returns on a missing handle instead of raising (#293) 3 people 2026-06-18 11:03:21 -05:00
• ec33b34279 docs: reconcile docs with shipped reality; fix Shell allowlist bug docs/review-fixes-and-shell-allowlist Aditya VardhanandClaude Opus 4.8 2026-06-18 15:20:12 +01:00
• 61d1d6cb0a fix(dynamic-workflow): address CodeRabbit review Aditya VardhanandClaude Opus 4.8 2026-06-17 15:54:32 +01:00
• 88720880e5 docs(dynamic-workflow): restructure README for faster comprehension Aditya VardhanandClaude Opus 4.8 2026-06-17 15:49:33 +01:00
• 583b762890 Merge branch 'main' into dynamic-workflow Aditya VardhanandClaude Opus 4.8 2026-06-17 15:40:09 +01:00
• e6dfd4c303 ci: add pydantic-ai v2 beta early-warning job and make code_mode v1/v2 compatible (#291) v0.4.0 4 people 2026-06-17 09:48:19 +01:00
• caf50108fe Explain code mode metadata selection (#292) David SFandGitHub 2026-06-16 18:49:53 -05:00
• 623a827128 Explain code mode metadata selection codex/describe-code-mode-metadata-selection David SF 2026-06-16 18:45:02 -05:00
• 24fd30e70c docs: stale-PR re-eval, experimental-export, and background-coder review guidance (#288) 4 people 2026-06-15 17:15:54 -05:00
• 2924277b17 Add CodeMode dynamic_catalog flag for cache-stable tool disclosure (#243) 4 people 2026-06-15 15:42:54 -06:00
• 9821aab0d5 feat(compaction): add ClampOversizedMessages for runaway generations (#286) 4 people 2026-06-15 15:34:19 -05:00
• 3ac6034f7e feat(experimental): OverflowingToolOutput capability for oversized tool returns (#290) 4 people 2026-06-15 13:11:53 -05:00
• 4bbb3e7431 refactor(overflow): convert remaining avoidable Any to precise types (A1) oversized-tool-output David SFandClaude Opus 4.8 2026-06-15 13:05:27 -05:00
• 2e4f9f3f38 fix(overflow): address review -- store hardening, content overflow, bounds, no-Any, TTL cleanup David SFandClaude Opus 4.8 2026-06-15 12:58:10 -05:00
• 5ce2e7af36 Merge main into capability/code-mode-dynamic-catalog capability/code-mode-dynamic-catalog David SFandClaude Opus 4.8 2026-06-15 12:44:01 -05:00
• c56a4b42f8 Add host-backed OS and filesystem access to CodeMode sandbox (#262) 4 people 2026-06-15 22:02:21 +05:30
• eb3201151a code_mode: warn default run_code prompt about unavailable fs/env; drop Any from CodeModeOSCallback claude/codemode-os-function-GK8ew David SFandClaude Opus 4.8 2026-06-15 11:31:18 -05:00
• 3942f63a27 feat(experimental): OverflowingToolOutput capability for oversized tool returns David SFandClaude Opus 4.8 2026-06-15 11:22:18 -05:00
• 0cfb8ee7ec code_mode: import TypedDict from typing_extensions for py<3.12 David SFandClaude Opus 4.8 2026-06-15 10:39:59 -05:00
• abaa14d9a1 code_mode: drop cast/Any from discovered-name extraction David SFandClaude Opus 4.8 2026-06-15 10:31:10 -05:00
• faef0e4889 Merge main into capability/code-mode-dynamic-catalog David SFandClaude Opus 4.8 2026-06-15 10:30:49 -05:00
• 9987ac7897 Add LocalStack capability for emulated AWS environments Claude 2026-06-04 04:51:02 +00:00
• cca154ce80 refactor: fix a comment stating a false invariant; unshadow a closure param Aditya VardhanandClaude Fable 5 2026-06-12 18:01:53 +05:30
• 64cea9703f refactor: narrow the shared executor seam and drop dead type unions Aditya VardhanandClaude Fable 5 2026-06-12 17:57:53 +05:30
• 2c3a281a06 refactor(dynamic-workflow): remove duplication that could drift Aditya VardhanandClaude Fable 5 2026-06-12 17:45:29 +05:30
• 96ad121ba0 fix(code_mode): convert sandbox VM panics to retries Aditya VardhanandClaude Fable 5 2026-06-12 17:15:40 +05:30
• 889628521b feat(dynamic-workflow): add a dedicated install extra Aditya VardhanandClaude Fable 5 2026-06-12 17:15:28 +05:30
• b225f47d6a docs(examples): replace the generated example with a feature-led README Aditya VardhanandClaude Fable 5 2026-06-12 17:15:27 +05:30
• a35aa05708 Merge origin/main into dynamic-workflow Aditya VardhanandClaude Fable 5 2026-06-12 17:00:39 +05:30
• 05ae737182 bumping pydantic ai (#280) 3 people 2026-06-12 16:55:22 +05:30
• 306a81049a Merge branch 'main' of github.com:pydantic/pydantic-ai-harness into dynamic-workflow Aditya Vardhan 2026-06-12 11:55:56 +05:30
• 3f49da59b4 docs(acp): appease codespell (unparseable -> unparsable) Aditya VardhanandClaude Fable 5 2026-06-12 11:21:03 +05:30
• d8d7046d5e Add Docling and Markdownify document-conversion capabilities claude/docling-markdownify-conversion-2o4e67 Claude 2026-06-12 05:50:59 +00:00
• a4d721c3b3 Merge branch 'main' of github.com:pydantic/pydantic-ai-harness into acp-support Aditya Vardhan 2026-06-12 11:10:27 +05:30
• ce7b1b3b97 refactor(acp): let the turn build its own response and drop the commit-flag tuple Aditya VardhanandClaude Fable 5 2026-06-12 11:06:28 +05:30
• e34c30ae54 feat(shell): control subprocess environment to stop secret inheritance (#284) 4 people 2026-06-11 22:51:41 -05:00
• 34dcbf8352 feat(subagents): per-delegate run controls (usage limits, timeout, call budget, soft-failure) (#283) 4 people 2026-06-11 22:50:11 -05:00
• 2d85cb38af review: validate limits keys, exclude run-state from eq, cover parallel budget subagents-run-controls David SFandClaude Opus 4.8 2026-06-11 22:47:28 -05:00
• 5087daa418 docs(shell): tighten env-control wording after adversarial review shell-env-control David SFandClaude Opus 4.8 2026-06-11 22:45:50 -05:00
• 1b0650a913 feat(compaction): add ClampOversizedMessages for runaway generations clamp-oversized-messages David SFandClaude Opus 4.8 2026-06-11 21:40:11 -05:00
• 40a60de696 feat(subagents): per-delegate run controls via SubAgentLimits David SFandClaude Opus 4.8 2026-06-11 13:48:35 -05:00
• 823ee3bc83 feat(shell): control subprocess environment to stop secret inheritance David SFandClaude Opus 4.8 2026-06-11 13:43:00 -05:00
• f10638569e docs(acp): record the post-commit-cancel semantics in CONFORMANCE Aditya VardhanandClaude Fable 5 2026-06-11 20:34:56 +05:30
• 6b25e59217 fix(acp): survive raw cancellation during terminal create; answer cancelled after a post-commit cancel Aditya VardhanandClaude Fable 5 2026-06-11 20:32:22 +05:30
• f79c547505 test(acp): pin claimed behavior the suite asserted too weakly or not at all Aditya VardhanandClaude Fable 5 2026-06-11 20:10:27 +05:30
• d96b6dab98 refactor(acp): small type and surface cleanups from review Aditya VardhanandClaude Fable 5 2026-06-11 20:05:06 +05:30
• 4682953718 refactor(acp): shield the terminal create in place instead of via a side task Aditya VardhanandClaude Fable 5 2026-06-11 20:05:02 +05:30
• 3086e625ef docs(acp): true up the conformance claims, comments, and doc links Aditya VardhanandClaude Fable 5 2026-06-11 19:53:54 +05:30
• f64c38449b fix(acp): resolve workspace-relative paths before they reach fs/read|write Aditya VardhanandClaude Fable 5 2026-06-11 19:51:25 +05:30
• 7fefa20a9e fix(acp): end limit-hit turns with max_turn_requests, close out tool calls on errors Aditya VardhanandClaude Fable 5 2026-06-11 19:49:15 +05:30
• cb1f85e12e fix(acp): don't leak a terminal when run_command is cancelled mid-create Aditya VardhanandClaude Fable 5 2026-06-11 19:44:24 +05:30
• 832470f800 docs(dynamic-workflow): record the core usage_limits TOCTOU race behind max_agent_calls Aditya VardhanandClaude Opus 4.8 2026-06-11 19:43:26 +05:30
• 4799e66604 fix(acp): don't report rollback when a cancel lands in the post-commit save Aditya VardhanandClaude Fable 5 2026-06-11 19:40:50 +05:30
• 4cc6e2ab63 fix(acp): never swallow a handler's own teardown cancellation Aditya VardhanandClaude Fable 5 2026-06-11 19:38:48 +05:30
• f4bf710b73 fix(acp): reject prompts queued behind a session that closed or reloaded Aditya VardhanandClaude Fable 5 2026-06-11 19:34:08 +05:30
• f888891866 test(dynamic-workflow): cover deps forwarding, message isolation, and shared-counter limits Aditya VardhanandClaude Opus 4.8 2026-06-11 19:18:56 +05:30
• 6f2aa11438 Add StepPersistence capability for step-event durability across delegates (#251) 4 people 2026-06-10 23:55:06 -05:00
• 6912941080 docs(experimental): add warning banner to planning/subagents, finish em-dash sweep step-persistence David SF 2026-06-10 23:52:57 -05:00
• 89372a2ea5 docs(media): fix metadata/key_strategy drift, convert em-dashes David SF 2026-06-10 23:44:34 -05:00
• e4906112d9 refactor: gate step_persistence + media behind experimental namespace David SF 2026-06-10 23:30:12 -05:00
• 0fc7ce9868 Merge remote-tracking branch 'origin/main' into step-persistence David SF 2026-06-10 23:21:18 -05:00
• ed7f687848 docs(dynamic-workflow): document experimental status and actual return shapes Aditya VardhanandClaude Opus 4.8 2026-06-10 15:32:24 +05:30
• 79fd37883e fix: await cancelled dispatch tasks and survive bad mid-run reveals Aditya VardhanandClaude Opus 4.8 2026-06-10 15:32:24 +05:30
• d347d08e3c feat(subagents): add SubAgents capability for delegating to child agents (#267) 3 people 2026-06-09 12:12:20 -04:00
• ae371bbdaf docs: ban em-dashes and codify writing style (#270) 3 people 2026-06-09 12:08:14 -04:00
• 3d3247110c feat(planning): add Planning capability with cache-safe plan reminders (#266) 3 people 2026-06-09 12:05:19 -04:00
• 4ce414461e fix(acp): close out in-flight tool calls when a turn is cancelled Aditya VardhanandClaude Opus 4.8 2026-06-09 19:50:44 +05:30
• 94d3c50e87 fix(dynamic-workflow): harden sandbox boundary and reveal/config validation Aditya VardhanandClaude Opus 4.8 2026-06-09 19:44:07 +05:30
• e5b7bf0c8b fix(acp): handle SessionStore failures instead of leaking them Aditya VardhanandClaude Opus 4.8 2026-06-09 18:56:15 +05:30
• 0e54b542b1 docs(dynamic-workflow): correct stdlib import guidance for the sandbox Aditya VardhanandClaude Opus 4.8 2026-06-09 18:50:07 +05:30
• 38ead88ef2 docs(acp): correct the SessionStore failure note to match observed behaviour Aditya VardhanandClaude Opus 4.8 2026-06-09 18:29:15 +05:30
• 89941b3541 docs(acp): record the unhandled SessionStore failure contract Aditya VardhanandClaude Opus 4.8 2026-06-09 18:20:36 +05:30
• 028ace63ca docs(acp): correct the API reference and capability notes Aditya VardhanandClaude Opus 4.8 2026-06-09 18:14:06 +05:30
• 84de0172ad fix(acp): reliable approval scope, mapped stop reasons, safe session reload Aditya VardhanandClaude Opus 4.8 2026-06-09 18:13:59 +05:30
• d6833de002 fix(acp): keep inline image bytes and let run_acp_stdio advertise MCP Aditya VardhanandClaude Opus 4.8 2026-06-09 18:13:51 +05:30
• 5744335df1 refactor(dynamic-workflow): move capability under experimental Aditya VardhanandClaude Opus 4.8 2026-06-09 17:51:24 +05:30
• 75474cacc8 Merge branch 'main' of github.com:pydantic/pydantic-ai-harness into dynamic-workflow Aditya Vardhan 2026-06-09 17:39:42 +05:30
• 6680a120dc refactor(acp): stop surfacing unconsumed additionalDirectories Aditya VardhanandClaude Opus 4.8 2026-06-09 15:23:49 +05:30
• 7f2e812bae test(acp): add spec-conformance suite driven over a real in-memory wire Aditya VardhanandClaude Opus 4.8 2026-06-09 15:04:45 +05:30
• d5ec3a959f docs(acp): record why _all_known_model_names avoids the unreleased public API Aditya VardhanandClaude Opus 4.8 2026-06-09 14:28:49 +05:30
• 450041ac02 fix(acp): make the collection guard coverable in a single environment Aditya VardhanandClaude Opus 4.8 2026-06-06 09:27:24 +05:30
• ad7046bc1a fix(acp): replay user turns on session/load, advertise MCP capabilities Aditya VardhanandClaude Opus 4.8 2026-06-06 09:13:34 +05:30
• c8b9441f71 docs(acp): point _all_known_model_names at the now-merged known_model_names() David SFandClaude Opus 4.8 2026-06-08 18:21:08 -05:00
• 4f4fde7e6a refactor(github): tighten after maintainer-lens review claude/exciting-turing-TmpC5 Claude 2026-06-08 02:27:44 +00:00
• f9e7c89c2f feat: add GitHub capability (GitHub MCP server in Docker) Claude 2026-06-08 02:09:22 +00:00
• d240a9363c Add ModalSandbox capability for isolated cloud sandboxes Claude 2026-06-04 04:50:40 +00:00
• 1ad812a514 feat(subagents): add SubAgents capability for delegating to child agents claude/subagents-capability Claude 2026-06-04 04:51:38 +00:00
• 82d5b95816 feat(planning): add Planning capability with cache-safe plan reminders claude/planning-capability Claude 2026-06-04 11:13:51 +00:00