tesserakit-license

Detect and classify a project's license, offline.

tessera-license reads LICENSE files and the license field of manifests, identifies the license by content signature, classifies it (permissive / weak-copyleft / copyleft / public-domain), and flags hygiene issues. No network and no per-dependency registry lookups: it inspects what is declared in the repo.

Audit

tessera license audit --input . --output ./out/license_pack

Detected sources:

• LICENSE / LICENCE / COPYING files (identified from their text)
• pyproject.toml [project].license
• package.json license
• Cargo.toml [package].license

Recognized ids include MIT, Apache-2.0, BSD-2/3-Clause, ISC, MPL-2.0, (L)GPL-2.0/3.0, AGPL-3.0, Unlicense, CC0-1.0.

Artifacts written:

licenses.jsonl           one LicenseFinding per declaration (id, category, evidence)
index.md                 the license inventory
validation_report.md     hygiene findings
coverage_report.md       counts by category and license id

Findings

• no_license — no LICENSE file or declared license found
• missing_license_file — a manifest declares a license but there is no LICENSE file
• copyleft_license — a copyleft license (GPL/AGPL) needs obligation review
• license_mismatch — different licenses declared in different places
• unrecognized_license — the license text/value could not be identified