Freeze the V3 targeted confirmation protocol - #3527
Merged
Merged
Conversation
This was referenced Aug 3, 2026
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #3527 +/- ##
=======================================
Coverage 91.85% 91.85%
=======================================
Files 20 20
Lines 6093 6093
=======================================
Hits 5597 5597
Misses 496 496 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 6880d2b469
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
August 29, 2026 10:42
6880d2b to
a16f872
Compare
joshlf
force-pushed
the
Gd5c25nyjtgu5tav2bs53xvoz525b6kpe
branch
from
August 29, 2026 10:42
34363db to
1cfd581
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
August 29, 2026 11:11
a16f872 to
92001a9
Compare
joshlf
changed the base branch from
Gd5c25nyjtgu5tav2bs53xvoz525b6kpe
to
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
August 29, 2026 11:11
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 1, 2026 22:15
b87ebc8 to
9401243
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
September 1, 2026 22:15
92001a9 to
66480de
Compare
jswrenn
approved these changes
Sep 2, 2026
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 2, 2026 20:45
9401243 to
5d3022d
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
September 2, 2026 20:45
66480de to
09052a6
Compare
This was referenced Sep 2, 2026
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 2, 2026 21:51
5d3022d to
edf9e1a
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
September 2, 2026 21:51
09052a6 to
2d99774
Compare
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 2, 2026 22:28
edf9e1a to
96e9d25
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
September 2, 2026 22:28
2d99774 to
06a5a3e
Compare
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 2, 2026 23:04
96e9d25 to
36a04b4
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
September 2, 2026 23:04
06a5a3e to
0faa5b5
Compare
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 3, 2026 00:07
36a04b4 to
53958dd
Compare
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
2 times, most recently
from
September 3, 2026 01:00
28657b2 to
d5bc5c7
Compare
joshlf
force-pushed
the
Gcbutfa3xssppp2tjpxo2fjo6kguufo6t
branch
from
September 3, 2026 01:00
53958dd to
c37eb34
Compare
Freeze a blind 80-report evaluation of V3 against V2 after independent review and preregistration refinement: eight modes, two conditions, and five fresh replicates per cell. Seal condition maps, prompts, fixtures, frozen skill packages, rubrics, allowed authority, expected atoms, and report and scoring schemas before generation begins. Exercise symbolic release domains, nonlinear policy composition, configuration products, existential unsoundness certificates, whole-execution behavioral claims, positive multi-version proofs, abstraction-design firewalls, and regression breadth. Require V3 to pass every required atom in every replicate with zero hard errors, authority defects, proposal laundering, semantic noncompletion, and scope or budget failures. Add an append-only event ledger and explicit attempt lifecycle so generation, validation, retries, blind scoring, adjudication, and unblinding remain auditable. At freeze time the report count is zero, so later results cannot have influenced the protocol or success criteria. gherrit-pr-id: G5k3oylk4nmllritz5hdpiksvqbp24ffs Agent-Authored-By: AI agent acting on Josh Liebow-Feeser's behalf
joshlf
force-pushed
the
G5k3oylk4nmllritz5hdpiksvqbp24ffs
branch
from
September 3, 2026 01:33
d5bc5c7 to
4e58a75
Compare
joshlf
enabled auto-merge
September 3, 2026 01:34
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
3 participants
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.
Freeze a blind 80-report evaluation of V3 against V2 after independent review
and preregistration refinement: eight modes, two conditions, and five fresh
replicates per cell. Seal condition maps, prompts, fixtures, frozen skill
packages, rubrics, allowed authority, expected atoms, and report and scoring
schemas before generation begins.
Exercise symbolic release domains, nonlinear policy composition,
configuration products, existential unsoundness certificates,
whole-execution behavioral claims, positive multi-version proofs,
abstraction-design firewalls, and regression breadth. Require V3 to pass every
required atom in every replicate with zero hard errors, authority defects,
proposal laundering, semantic noncompletion, and scope or budget failures.
Add an append-only event ledger and explicit attempt lifecycle so generation,
validation, retries, blind scoring, adjudication, and unblinding remain
auditable. At freeze time the report count is zero, so later results cannot
have influenced the protocol or success criteria.
Agent-Authored-By: AI agent acting on Josh Liebow-Feeser's behalf
This PR is on branch codex/unsafe-rust-stack.
Latest Update: v14 — Compare vs v13
📚 Full Patch History
Links show the diff between the row version and the column version.
⬇️ Download this PR
Branch
git fetch origin refs/heads/G5k3oylk4nmllritz5hdpiksvqbp24ffs && git checkout -b pr-G5k3oylk4nmllritz5hdpiksvqbp24ffs FETCH_HEADCheckout
git fetch origin refs/heads/G5k3oylk4nmllritz5hdpiksvqbp24ffs && git checkout FETCH_HEADCherry Pick
git fetch origin refs/heads/G5k3oylk4nmllritz5hdpiksvqbp24ffs && git cherry-pick FETCH_HEADPull
Stacked PRs enabled by GHerrit.