Skip to content

docs: bound SafeMini execution claims - #5

Merged
JustinJLeopard merged 1 commit into
mainfrom
desktop-codex/docs/safe-mini-execution-claims
Aug 10, 2026
Merged

docs: bound SafeMini execution claims#5
JustinJLeopard merged 1 commit into
mainfrom
desktop-codex/docs/safe-mini-execution-claims

Conversation

@JustinJLeopard

Copy link
Copy Markdown
Owner

Summary

Align source documentation and package metadata with the actual executor boundary.

  • remove standalone sandbox claims for host-process policy executors
  • remove unsupported incident artifact, trajectory, and ledger claims
  • distinguish in-memory run transcripts from persisted artifacts
  • preserve explicit injected-executor isolation as the supported boundary

Validation

  • git diff --check
  • PYTHONDONTWRITEBYTECODE=1 /usr/bin/python3 -B -m pytest -q -p no:cacheprovider (55 passed)

Non-goals

No executor behavior, runtime activation, package release, tag, or publication change.

Describe host-process policy guards and in-memory transcripts accurately. Preserve the distinction between explicit executor isolation and the built-in runner.

Co-Authored-By: Codex <noreply@openai.com>
Copilot AI lite review requested due to automatic review settings August 10, 2026 13:54
@JustinJLeopard
JustinJLeopard merged commit 9cbb5eb into main Aug 10, 2026
3 checks passed
@JustinJLeopard
JustinJLeopard deleted the desktop-codex/docs/safe-mini-execution-claims branch August 10, 2026 13:55

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR tightens and corrects SafeMini’s documentation and public metadata to match the actual execution/isolation boundary: built-in executors run in the host process context with policy guards, while isolation is only guaranteed when an injected executor boundary is used.

Changes:

  • Updated type/protocol docstrings to describe RunResult as carrying an in-memory transcript (not a persisted trajectory/audit artifact).
  • Revised README claims to remove standalone sandbox/incident/ledger assertions and clarify the supported isolation boundary.
  • Adjusted package description/keywords in pyproject.toml to match the bounded, policy-guarded execution model.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated no comments.

Show a summary per file
File Description
safe_mini/types.py Updates RunResult docstring to reflect in-memory transcript semantics.
safe_mini/protocol.py Reframes AgentRunner contract language from “sandboxed” to “policy-guarded” and aligns return expectations with transcript-based results.
safe_mini/policies/executor.py Narrows SafeExecutor docstring to “restrictive policy” wording consistent with host-context execution.
README.md Removes/clarifies overstated isolation and artifact claims; documents injected-executor boundary as the isolation guarantee.
pyproject.toml Aligns package description/keywords with bounded local execution + policy guards.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants