Skip to content

Releases

Suno for CrewAI 0.1.0

Choose a tag to compare

@acedatacloud-dev acedatacloud-dev released this 18 Oct 23:02
5b3bc4d

Suno by Ace Data Cloud for CrewAI — 0.1.0

Stable Ace repository release, built from merged main commit 5b3bc4d597f5fb046b1d6298d5e833ad7f2b29fb. Source PR #1 is MERGED. This is not official CrewAI default inclusion, an AMP listing, or a PyPI publication.

Normal wheel installation (downloaded artifact must match the digest):

uv add 'crewai-suno @ https://github.com/AceDataCloud/SunoCrewAI/releases/download/v0.1.0/crewai_suno-0.1.0-py3-none-any.whl#sha256=c00d2d011c72f0b1d294ce020b06c20668ef8d4572aaffbc53874341bc707ecf'

Source install:

uv add 'crewai-suno @ git+https://github.com/AceDataCloud/SunoCrewAI.git@5b3bc4d597f5fb046b1d6298d5e833ad7f2b29fb'

The tag/release page is informational; use immutable source or SHA256 as the trust anchor. Digests:

  • crewai_suno-0.1.0.tar.gz: sha256:675121850a37e87d92d360b43499a469c0b0990f6b07d21371049391729a9252
  • crewai_suno-0.1.0-py3-none-any.whl: sha256:c00d2d011c72f0b1d294ce020b06c20668ef8d4572aaffbc53874341bc707ecf

Native CrewAI generate and original-task retrieve tools, explicit service schemas, API-consumer Key from environment, single asynchronous submission without paid HTTP retry/fallback, bounded same-ID polling. Explicit failure is terminal even before finished_at; empty terminal media is not success. HTTP status is retained without raw service-body/credential exposure; result cost fields are filtered.

Complete Linux Python3.10/3.12/3.13 installation, standard test discovery, lint, build and clean wheel schema/import verified on exact source via CI. Tests use local deterministic model fixtures and mock paid generation. A historical task was queried without regenerating; no new/unique-bill, paid model decision, current media-playback or revenue claim.

Generation and any LLM configured by users remain billable at their current prices. Check current service/model permissions, balance and Token restrictions. Never place credentials in prompts, shared logs or screenshots. English/Chinese onboarding is now available on main with authentic redacted Key images and historical official UI references, not claimed live tool screenshots.

PyPI new-project publishing uses the existing publisher identity and normal shared quota; no duplicate account or bypass. Official Provider and five-service Tools submissions remain separately reviewed by CrewAI upstream.

Suno for CrewAI 0.1.0 review preview

Pre-release

Choose a tag to compare

@acedatacloud-dev acedatacloud-dev released this 18 Oct 05:17

Review preview of Suno by Ace Data Cloud for CrewAI. This is an Ace repository prerelease, not a PyPI release or an official CrewAI catalog entry. Implementation PR remains open: #1 . No merge has been performed.

Native Generate Music / Retrieve Task tools, explicit input schemas, single asynchronous generation attempt, same-ID bounded retrieval, and structured task failures. Includes English/Chinese Key onboarding with redacted authentic screenshots and four supported source tags.

Linux Python 3.10 / 3.12 / 3.13 complete dependency install, 14 no-bill tests, lint and wheel/sdist build passed in CI, including an actual CrewAI Agent/Task/Crew cycle with a deterministic local model fixture. This does not validate real paid model decisions or new media generation. One historical Suno task was queried without regenerating (two audio results, original cost 0.504 Credits).

Install the wheel only with its digest pinned: uv add 'crewai-suno @ https://github.com/AceDataCloud/SunoCrewAI/releases/download/v0.1.0-preview.1/crewai_suno-0.1.0-py3-none-any.whl#sha256=d31998ca7f33f54c9603a0acb0d906f25933e858be6c83babc5a6d0e5a478e51'. The source archive SHA256 is 515f7d600b45aeed069cb4fae331bcfeedf790f421f7fa8b0d51f71ec2530429. The wheel and source archive were built from commit e5eeb04661ed7b9a86e1f2f970a9e7943b027ca2; do not treat a mutable release tag as the trust anchor. API generation and any LLM configured by the user are billable at their current prices. No automatic paid HTTP retry or model fallback. Accepted/pending is not complete; retain and query the same task ID. Intel macOS may lack the mandatory lancedb dependency wheel; use supported Linux rather than skipping dependencies for production.