docs: record qualified codec 0.2.0 publication

[tamashi095]

Summary

Records the codec 0.2.0 publication, as 10abab9 did for 0.1.1: a "Published release" paragraph in .github/ISSUE_SPECS/010-move-the-codec-to-stable-effect-4-0-x.md and .github/RELEASE_EVIDENCE/010/registry-release.json (version, registry integrity, tarball SHA-256, source revision, decoder SHA-256, both workflow runs).

• Publish run https://github.com/misofm/codec/actions/runs/38429325908 : qualify and npm publish (signed provenance) succeeded; the post-publish convergence check failed only because the registry took more than two minutes to list the version, so registry-consumer was skipped.
• Verify run https://github.com/misofm/codec/actions/runs/38430320099 at the same commit: qualify, publish-or-compare and registry-consumer all succeeded against the published tarball.
• Tag/release v0.2.0 (annotated tag at 56383ab): https://github.com/misofm/codec/releases/tag/v0.2.0
• Independently re-derived from the registry tarball: SHA-256 dcc05849…e999, SHA-512 integrity matches, packaged peer effect >=4.0.3 <4.1, decoder asset 70caf381….

Docs/evidence only; no source change. The throwaway effect-4.0.4-ci-dist branch is intentionally not deleted yet (the adapter PR #130 history still references it).

Closes #10.

🤖 Generated with Claude Code