Hunter Bown 401c1f6cf8 fix(security): tighten approval policy for run_tests ...

`run_tests` declares `ToolCapability::ExecutesCode` and runs `cargo test`,
which executes workspace code at test, build-script, and proc-macro time.
Match the default approval policy for code-executing tools by returning
`ApprovalRequirement::Required` explicitly instead of overriding to
`Auto`. Adds a regression test pinning the approval requirement.

2026-05-08 18:58:01 -05:00

..

assets/skills/skill-creator

feat(skills): #139 bundle skill-creator and auto-install on first launch

2026-04-27 23:52:58 -05:00

src

fix(security): tighten approval policy for run_tests

2026-05-08 18:58:01 -05:00

tests

fix(skills): accept workflow pack archive layouts (#1164)

2026-05-08 02:37:21 -05:00

build.rs

fix(provider): enable OpenAI-compatible TUI runtime (#1017)

2026-05-07 05:32:15 -05:00

Cargo.toml

chore(release): prepare v0.8.23

2026-05-08 18:17:44 -05:00