Skip to content

feat(implement): hardening asks in plain words; boundary validation is never speculative - #26

Merged
leomontigatti merged 1 commit into
mainfrom
feat/plain-language-hardening-asks
Oct 23, 2026
Merged

leomontigatti merged 1 commit into
mainfrom
feat/plain-language-hardening-asks

Conversation

@leomontigatti

Copy link
Copy Markdown
Owner

This changes what every project's agent does in two ways. A speculative hardening ask is now written in plain words someone outside the code can rule on: what breaks and for whom, how likely it is, and what the fix costs. Validating input where it enters from outside the system (a user, the network, another service) is never speculative, so the agent adds it without asking. No project has to add or rename anything.

The definition stays in `implement` step 2. The implementer brief's `risks` line now asks for the same plain-words shape. Building on #22. The idea comes from a public post, and no text is carried over, so there's no `decisions.md` row and no `NOTICE` entry.

Review

Tier: readback, reviewed 87b0920

  • readback (Luna): 1 finding, declined (the implementer brief points to step 2 for the definition and does not repeat its exceptions, including the existing `critical` one)

🤖 Generated with Claude Code

…s never speculative

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@leomontigatti
leomontigatti merged commit 75d67d0 into main Oct 23, 2026
3 checks passed
@leomontigatti
leomontigatti deleted the feat/plain-language-hardening-asks branch October 26, 2026 15:35
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.

1 participant