Item detail

github.com

openai/openai-agents-python

openai/openai-agents-python is a framework in RepoRadar's Developer Tools section, holding Gold tier and a 'try now' verdict. Its strongest signal is workflow potential, scored 10.0 out of 10.

Open on GitHub →Back to radar

Score8.9

Popularity84.0

Riskconditional

TierGold

Score breakdown

Usefulness9.0

Novelty7.0

Momentum8.0

Maturity8.6

Open-source/build8.4

Evidence7.2

Workflow potential10.0

Setup ease6.4

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for Python teams that want a first-party agent runtime with tracing and guardrails instead of stitching together tool use, orchestration, and review logic by hand.

Who should use it

Python developers teams building agent products automation engineers voice and tool-using agent builders

Who should skip it

Skip openai/openai-agents-python if the source repository or demo is inactive, unmaintained, or no longer matches the description shown here.

About this signal

openai/openai-agents-python is tracked by RepoRadar as a framework in the Developer Tools section. First seen 2026-06-20; the source record was last checked on 2026-06-20. The current verdict is 'try now' with a Gold tier and moderate setup difficulty. The standout signals for openai/openai-agents-python are workflow potential (10.0) and practical usefulness (9.0), while setup ease (6.4) trails — that balance shapes where it fits best. This page summarizes the public evidence on the linked source page and states where additional review is still needed.

How this item is evaluated

The openai/openai-agents-python record combines a 8.9/10 composite score with separate popularity (84.0), risk (conditional), and setup (moderate) signals. See the scoring methodology for the current weights and evidence definitions.

Putting this into practice? Read How to vet an AI agent or MCP server before you wire it in for the checklist behind this score.

Risk explanation

The SDK is built to broker tool calls, MCP servers, sandbox execution, and voice or realtime flows into external systems, so start with least-privilege credentials and explicit approval on any side-effecting actions.

Evidence links

github.com

Closest alternatives / related signals

openai python agent-sdk mcp voice-agents

Verification record

What RepoRadar actually verified

Tested in a bounded workflow

Bounded representative workflow retained by RepoRadar verification harness. Last checked 2026-07-13T10:35:57.333092Z.

passed · cohort-20260712-openai-agents-construct

Tester: RepoRadar automated local verification harness
Started: 2026-07-13T03:22:36.270448Z
Completed: 2026-07-13T03:22:40.567152Z
Environment: Windows 10 AMD64; Python 3.11.9; credential-stripped child environment; disposable home/cache
Install/setup time: 0 minute(s)
Evidence scope: Bounded setup or capability check
Cleanup: Per-check temporary home and work directory removed. Shared cohort package cache removed.

Actions exercised

Created a disposable home, work directory, and isolated package cache with credential-like environment variables excluded.
Executed bounded check: Construct an OpenAI Agents SDK agent without running it.
Captured the complete sanitized stdout, stderr, exit status, and 4.30-second wall time.

Observed results

Command exited 0 after 4.30 seconds.
Expected marker 'CHECK_OK' was observed in retained output.

Observed strengths

The OpenAI Agents SDK constructed a named Agent with local instructions without invoking a runner, tool, trace exporter, credential, or model.

Friction

The command emitted stderr; warnings or errors are preserved in the retained log for review.

Limitations

No runner, tool, trace export, API credential, or model call was used.
This bounded cohort check is not a production benchmark or a claim of real user-workflow adoption.

Pricing assessment: No paid plan or metered provider usage was exercised; package or licensing, hosting, and provider costs remain workflow-dependent.

Privacy assessment: No repository content, user data, or provider prompt was transmitted; broader product data handling was not assessed by this bounded run.

Open retained test log →

passed · cohort-20260712-openai-agents-function-tool-workflow

Tester: RepoRadar automated local verification harness
Started: 2026-07-13T10:35:52.923510Z
Completed: 2026-07-13T10:35:57.333092Z
Environment: Windows 10 AMD64; Python 3.11.9; credential-stripped child environment; disposable home/cache
Install/setup time: 1 minute(s)
Evidence scope: Bounded representative workflow
Cleanup: Per-check temporary home and work directory removed. Shared cohort package cache removed.

Actions exercised

Created a disposable home, work directory, and isolated package cache with credential-like environment variables excluded.
Created 1 synthetic fixture file(s) inside the disposable work directory; retained hashes prove the exact inputs.
Decorated a typed Python function with the SDK's function_tool adapter and inspected its generated JSON schema.
Created a local ToolContext, dispatched serialized severity input through on_invoke_tool, and retained the schema and page result.
Executed bounded check: Generate and invoke a typed OpenAI Agents SDK function tool entirely locally.
Captured the complete sanitized stdout, stderr, exit status, artifact checks, and 4.41-second wall time.

Observed results

Command exited 0 after 4.41 seconds.
The SDK generated an integer-typed tool schema, parsed the JSON argument, invoked the function, and returned page.
Expected marker 'CHECK_OK tool=triage type=integer result=page' was observed in retained output.
Validated result.json: 3 required marker(s) present and 0 excluded marker(s) absent; size and SHA-256 are retained.

Observed strengths

Function-tool adaptation, strict schema generation, JSON argument handling, and invocation were usable independently of an API-backed agent run.

Friction

Direct invocation requires constructing a ToolContext; the full Runner normally supplies this context during a model-selected tool call.
Setup or runtime emitted 1 stderr line(s); the complete warnings/errors are preserved in the retained log.

Limitations

Direct function-tool invocation validates schema generation and input dispatch, not an Agent Runner, model selection, handoffs, tracing export, or hosted tool calls.
This credential-free disposable workflow does not establish production scale, model quality, reliability under sustained use, or team adoption.

Pricing assessment: The typed tool workflow used the local open-source SDK and made no OpenAI API, model, or hosted tracing request.

Privacy assessment: Only the synthetic severity argument was parsed in process; no API key, prompt, trace, or network transport was present.

Open retained test log →

Verification sources

Official source ↗ · checked 2026-08-13

Longitudinal intelligence

How this decision record is moving

Raw history JSON →

45 dated snapshots retained from 2026-06-20 through 2026-08-13; see the snapshot index for explicit coverage gaps. Stars, version, release, pricing, integration, risk, maintenance, verdict, score, and momentum fields remain explicit even when a source has not reported them. Repository momentum is a normalized 0–10 RepoRadar signal; GitHub stars appear only where the popularity monitor retained exact timestamped observations.

RepoRadar score8.9 current · +0.0 net

Repository momentum9.6 current · +1.6 net

GitHub stars (observed)28,617 current · +727 net

GitHub stars28,617 exact observation

Versionv0.20.0

Last release2026-08-11T03:12:14Z

Maintenanceactive

Current riskconditional

Current verdicttry now

Pricing baselineNo structured commercial pricing baseline

Pricing checkedNot applicable or not recorded

Pricing freshnessNo dated commercial pricing review

Integrations baselineNo structured integrations recorded

Recent dated points

Date	Score	Momentum	Stars	Risk	Verdict	Maintenance
2026-08-13	8.9	9.6	28,617	conditional	try now	active
2026-08-12	8.9	9.6	28,578	conditional	try now	active
2026-08-11	8.9	9.6	28,555	conditional	try now	active
2026-08-10	8.9	9.6	28,533	conditional	try now	active
2026-08-09	8.9	9.6	28,501	conditional	try now	active
2026-08-08	8.9	9.6	28,489	conditional	try now	active
2026-08-07	8.9	9.6	28,351	conditional	try now	active
2026-08-06	8.9	8.0	Not recorded	conditional	try now	not recorded
2026-08-05	8.9	8.0	Not recorded	conditional	try now	not recorded
2026-08-04	8.9	9.6	28,351	conditional	try now	active
2026-08-03	8.9	9.6	28,351	conditional	try now	active
2026-08-02	8.9	9.6	28,340	conditional	try now	active

Why the record changed

stars changed2026-08-13

Stars changed: 28578 → 28617.

stars changed2026-08-12

Stars changed: 28555 → 28578.

version changed2026-08-11

Version changed: v0.19.4 → v0.20.0.

stars changed2026-08-11

Stars changed: 28533 → 28555.

stars changed2026-08-10

Stars changed: 28501 → 28533.

stars changed2026-08-09

Stars changed: 28489 → 28501.

version changed2026-08-08

Version changed: v0.19.2 → v0.19.4.

stars changed2026-08-08

Stars changed: 28351 → 28489.

stars changed2026-08-03

Stars changed: 28340 → 28351.

stars changed2026-08-02

Stars changed: 28325 → 28340.

stars changed2026-08-01

Source-observed stars changed: 28321 → 28325. This reports the retained observation delta and does not infer why the upstream change occurred.

version changed2026-08-01

Source-observed version changed: v0.19.1 → v0.19.2. Associated release timestamp: 2026-07-29T07:44:24Z → 2026-08-01T01:08:37Z. This reports the retained observation delta and does not infer why the upstream change occurred.