Skip to content

Releases: xwh-01/dsh-mediacrawler

dsh-mediacrawler v0.3.0

Choose a tag to compare

@xwh-01 xwh-01 released this 14 Aug 09:10

Security and lifecycle hardening

  • Disable comments by default; collection now requires explicit comment opt-in.
  • Redact MediaCrawler platform credentials including msToken, a_bogus, verifyFp, verifyUuid, webid, and X-Bogus.
  • Rename exports as credential-redacted and explicitly report that PII is not anonymized and archives are not automatically safe to share.
  • Add an incremental append-aware JSONL index and remove duplicate result scans.
  • Add a 256 MiB default export limit, cancellation-safe workers, temporary-file cleanup, and cross-process export locking.
  • Add confirmed single-run deletion and dry-run-first retention cleanup without deleting browser login profiles.
  • Expand the MCP surface to 12 tools.

Verification

  • 96 Python tests and Node Skill provider test
  • Real MediaCrawler CLI contract and system Chrome probe
  • Clean public Git install and real DSH Web + MCP startup
  • Windows/Linux, Python 3.11/3.13, Node 22/24 CI

dsh-mediacrawler v0.2.0

Choose a tag to compare

@xwh-01 xwh-01 released this 14 Aug 08:28

What changed

  • Add runs to recover recent durable run IDs after restart or context loss.
  • Add result to return lifecycle status, typed artifacts, and a bounded redacted sample in one call.
  • Update the packaged Skill to use the shorter recovery and result workflow.
  • Clarify when MediaCrawler collection adds value beyond Harness web search providers.

Verification

  • 76 Python tests
  • Node Skill provider test
  • Python wheel and sdist builds
  • DSH bundle installation and real Web + MCP startup
  • Windows/Linux, Python 3.11/3.13, Node 22/24 CI

dsh-mediacrawler v0.1.0

Choose a tag to compare

@xwh-01 xwh-01 released this 14 Aug 08:02

dsh-mediacrawler v0.1.0

First independent ecosystem release of the installable DeepSeek Harness profile bundle for MediaCrawler.

Highlights

  • Installs into a DSH profile with dsh plugin --profile web add github:xwh-01/dsh-mediacrawler#v0.1.0.
  • Exposes eight bounded MCP tools for checks, collection, supervision, previews, and sanitized exports.
  • Ships the mediacrawler-collector Skill through a package-owned provider.
  • Keeps the adapter and MediaCrawler runtimes separate and uses isolated Chrome profiles by default.
  • Redacts logs, previews, manifests, and exported ZIP files.

Compatibility

  • DeepSeek Harness 0.1.0-rc.6
  • Node.js 22.19+ on the 22.x line, or Node.js 24+
  • Python 3.11 and 3.13
  • MediaCrawler command contract at 5665a271ef15e0ec82b1f48a951b66760e054db9

SHA-256

  • dsh-mediacrawler-0.1.0.tgz: 5DF27A5DD690049C1CFEBFDC057B6F1135CA0824021C8E2AC354C53EED3BEC34
  • dsh_mediacrawler-0.1.0-py3-none-any.whl: 06A67DA357A2BB9AAE7CC8CB36D0808CE48D2CB0385D78DE5D57A6799D957A56
  • dsh_mediacrawler-0.1.0.tar.gz: F448F47CAEEFB0B6D46A0BF197B335CC475A0242AB8D674515113C073483995B

MediaCrawler remains a separate project under its own non-commercial learning license and usage restrictions.