Closed
Conversation
agents should verify what they're about to execute before handing over credentials. this adds a non-negotiable audit step with concrete checks: no install scripts, no unexpected outbound calls, key material stays local, bridge script does what it says. applies to this skill specifically but the pattern should be standard for any skill that asks an agent to run third-party code.
ApprovabilityVerdict: Needs human review Unable to check for correctness in dd6152a. You can customize Macroscope's approvability policy. Learn more. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
adds a step 0 to the xmtp-agent skill: audit the code before running anything.
agents are autonomous. when you hand them a skill that says 'run this,' the first thing they should do is verify it deserves that trust. this adds concrete checks before install or init:
marked as non-negotiable because it should be. this pattern isn't specific to the xmtp-agent skill either, it should be standard for anything that asks an agent to execute third-party code with credentials.
no functional changes to the skill itself. just shifts step 1 to step 1 and puts the audit in front of it.
Note
Add pre-run security audit checklist to xmtp-agent skill documentation
Adds a "Step 0: Audit Before You Run" section to SKILL.md that instructs users to inspect npm scripts, grep for unexpected outbound network calls in shell/JS/Python files, and verify that key material under
~/.xmtpis not exfiltrated before executing any skill. Includes a checklist of pass/fail criteria and advises users to stop if any check fails.Macroscope summarized dd6152a.