chore(release): 0.8.1

[krassx]

Patch release 0.8.1: version bump (Cargo.toml, Cargo.lock) and the changelog entry for #79 (--type elf upgrades a stored symbol table to DWARF without --force, closes #78).

After merge: tag v0.8.1 -> release.yml (cargo-dist) -> npm-publish.yml + mirror-to-s3.yml chain off it via workflow_run.

🤖 Generated with Claude Code

[cursor]

Deep review — chore(release): 0.8.1

Reviewed the release plumbing in this diff and the #79 ELF upgrade path it actually ships (fe5caaf: format_variant / replace_if_richer on --type elf). Did not treat the three-file bump as self-contained.

Findings

No P0–P3 issues. Nothing that would mis-tag, mis-announce, or change upload/exit behavior relative to what #79 already merged.

What was checked

• Scope. v0.8.0 is already the GitHub latest. The only commit since then is #79. Changelog covers that and nothing else. [Unreleased] stays empty.
• Version consistency. Cargo.toml, Cargo.lock (bugsee-cli), and cargo metadata --no-deps are all 0.8.1. env!("CARGO_PKG_VERSION") (update, User-Agent) picks it up. npm templates stay 0.0.0-dev and are stamped at publish.
• Semver. Additive wire fields (skip_serializing_if none), no exit-code change. Patch is the same rule as 0.7.12 / 0.7.13; 0.8.0 was a minor because it did change exit codes.
• Notes vs code. ELF-only POST of format_variant + replace_if_richer: true; other flows still omit both; --type rust ELF still omits both (documented); classification is from has_debug_info / has_symbols, not the filename; --force still sends overwrite plus the variant; skip advice is gated on variant == dwarf && stored.is_none(), so a current server that reports params.format_variant does not print --force. .dynsym → symtab missed-upgrade is documented.
• cargo-dist. Plan job: announcement_tag v0.8.1, title 0.8.1 - 2026-10-08, changelog body is exactly the new section (including the #78 footnote). Empty Unreleased did not leak.
• CI. Green: rustfmt, clippy, test (ubuntu/macos/windows/windows-11-arm), e2e (all four), npm scripts, install.ps1, win-arm64 release build, dist plan. Artifact jobs correctly skipped on the PR.

Residual (not blocking)

The notes already say integrators should not adopt this until appserver #65–#67 is in production. Publishing the CLI first is still safe: extra JSON fields are ignored on an old server, and a skipped FULL library still gets the --force hint. bugsee-cli update (same major) will pick 0.8.1 up once mirrored — that is the intended two-step floor rollout, not a reason to hold the tag.

---

1. Overall risk: Low
2. Merge recommendation: Approve. Same release shape as #77. Tag v0.8.1 only after merge, not from this PR.
3. Most important issues to fix: None.
4. Positives: Changelog is unusually precise for a patch (wire fields, old-server fallback, known limits). Dist announcement matches. Tests around #79 pin the request body, the upgrade log, and the --force advice split.

  

_{Sent by Cursor Automation: Bugsee code review}