ABI Layer 5: end-to-end soundness capstone certificate#41
Merged
Conversation
Add Mylangiser.ABI.Semantics, a machine-checked flagship proof for the
headline domain property: progressive-disclosure interfaces only ever ADD
revealed items across steps, never remove a previously-shown one.
Model: a disclosure run is a sequence of revealed-item sets (List (List Nat)).
A valid step Reveal before after exists only when Subset before after holds
(every item visible before is still visible after) — there is no constructor
that drops a revealed item. MonotoneRun is the transitive closure over steps.
Includes a sound+complete decision procedure (decSubset, decReveal, decRun),
a certifier into the ABI Result type with a soundness lemma (certifyRunSound),
a positive control (goodRunMonotone: {1}->{1,2}->{1,2,3}) and a negative
control (hidingStepInvalid / hidingRunNotMonotone: {1,2}->{1} has no valid
transition proof). No believe_me/postulate/assert — fully total.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
Adds a second, deeper, distinct machine-checked theorem over the existing Semantics model. Where Layer 2 proves single-step monotonicity (one Reveal step may only add items), Layer 3 proves the multi-step disclosure relation Reaches is transitive (reachesTrans) and that reachability implies a global Subset (reachesImpliesSubset) — revealed sets only grow across many steps, a fact not implied by the local theorem. Includes the load-bearing algebraic laws subsetElem/subsetTrans/subsetRefl, a sound+complete decision procedure, a Result certifier with soundness extraction, a positive control (concrete two-hop reachability witness whose growth is machine-derived), a composed-path witness via transitivity, and negative/non-vacuity controls (hidingNotSubset, hidingNotReachable, hidingCertifyRejected). Builds clean with zero warnings; adversarial false claims are rejected by the checker. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
Seals the proof side of the ABI<->FFI seam: the FFI result-code encoding resultToInt : Result -> Bits32 is proven SOUND. - intToResult decoder + resultRoundTrip: encoding is lossless/faithful - resultToIntInjective: distinct ABI outcomes never collide on the wire, derived from the round-trip via cong + Just-injectivity - same round-trip + injectivity for disclosureLevelToInt (the other FFI enum encoder in this ABI) - positive controls (concrete decode = Refl) and a machine-checked non-vacuity control (distinct codes have distinct ints) Genuine proof only: no believe_me/idris_crash/assert_total/postulate. %default total throughout. Builds clean with zero warnings; a false seam claim (resultToInt Ok = resultToInt Error) is rejected by the checker. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
Assemble the existing Layer-2/3/4 ABI proofs into one inhabited end-to-end certificate, Mylangiser.ABI.Capstone: - ABISound record whose fields are the key proven ABI facts: flagship single-step monotonicity (Semantics.goodRunMonotone), the Layer-3 global invariant (Invariants.goodReachGrows), and the FFI-seam injectivity theorem (FfiSeam.resultToIntInjective). - abiContractDischarged : ABISound, the single inhabitant built entirely from those real exported witnesses. It type-checks iff every prior layer is sound, tying manifest -> ABI -> FFI seam into one soundness statement. Genuine composition only: no believe_me / idris_crash / assert_total / postulate / sorry. Adversarial check confirmed a bogus globalInvariant field (Subset [2] [1,2,3] vs required Subset [1] [1,2,3]) is rejected. Build is clean with zero warnings; %default total; SPDX header. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
Signed-off-by: Jonathan D.A. Jewell <6759885+hyperpolymath@users.noreply.github.com>
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
2 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.
Summary
Layer 5 (the capstone, completing the 5-layer proof track): a new
*.ABI.Capstonemodule importing every prior layer and assembling a single inhabitedABISoundcertificate (abiContractDischarged) from the real exported witnesses of the flagship property (L2 disclosure monotonicity), the deeper invariant (L3 transitivity), and the FFI-seam injectivity (L4). One end-to-end soundness statement.Genuine composition only — reuses real exported names.
Testing
Idris2 0.7.0
--build→ exit 0, zero warnings. Adversarial: a bogus-field certificate was rejected.build/removed. Nobelieve_me/postulate/sorry.🤖 Generated with Claude Code
https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
Generated by Claude Code