Skip to content

Honest agent wording: deterministic -> constrained in system prompt#1

Merged
dolepee merged 1 commit into
masterfrom
fix/constrained-agent-prompt
Jun 4, 2026
Merged

Honest agent wording: deterministic -> constrained in system prompt#1
dolepee merged 1 commit into
masterfrom
fix/constrained-agent-prompt

Conversation

@dolepee

@dolepee dolepee commented Jun 4, 2026

Copy link
Copy Markdown
Owner

The README deterministic wording was already fixed (the submission-framing polish). This catches the one that was missed: the agent system prompt at script/RequestHelloDecision.s.sol:19 still said "a deterministic DAO voting agent."

LLM output is not deterministic; the agent is constrained to the YES/NO/ABSTAIN allowed-value schema. Changed deterministic -> constrained to match what the code actually enforces and to remove the overclaim a judge would catch when reading the code.

String literal only: no contract change, no redeploy, no effect on the existing onchain proofs (already recorded). The third-party lib/forge-std uses of "deterministic" are unrelated and untouched.

The README was already clean, but the base-decision script's LLM system prompt
still called Steward a 'deterministic DAO voting agent'. LLM output is not
deterministic; the agent is constrained to the YES/NO/ABSTAIN schema. Corrects
the wording per the audit. String literal only: no contract change, no redeploy,
existing onchain proofs unaffected.
@dolepee dolepee merged commit 530e0bb into master Jun 4, 2026
6 checks passed
@dolepee dolepee deleted the fix/constrained-agent-prompt branch June 4, 2026 11:27
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