agent: builder model: opus max_turns: 100 budget_usd: 20 —
.bounds expectation beside a test program, so a proof is guarded by a testFacts: Where things live, Traps already met, Commands
Score: S=3 (every prover gain of the lane, p2-p4 and the counters next, is guarded today by no test: a later change can lose a proof silently, and claim (b)’s figure moves only when someone measures), O=3 M (every later brief of the lane tests with it; CLAUDE.md wants tests in Nikaia, and tests/compiler cannot show --bounds), C=6 M (one optional file pair in project_tests.nika, the comparison in the program-test runner, bless). Area: prover, toolchain.
tests/compiler is one project: a test cannot set --optimization and cannot show the --bounds report, so p2-p4 verified their gains with probes outside the suite (docs/briefs/prover/state.md, Lane facts). nikaia.toml has no optimization key. check --bounds FILE prints a first line bounds in FILE: N indexes, P proved; M operations, Q proved - written without their check: indexes off, operations off, then one line per site ( LINE TEXT proved|checked: REASON).
A program test tests/NAME.nika (ADR-269 D22) may have, beside its .stdout/.stderr:
NAME.bounds: the --bounds report of NAME.nika at prove-through-loops:off;NAME.loops.bounds: the same at prove-through-loops:on.nikaia test compares each one that exists, byte for byte, after two normalisations: the path in the first line is shown relative to the project (tests/NAME.nika), and the build-dependent tail ( - written without their check: ...) is cut, so the file depends neither on --both-settings nor on the machine. A difference fails that test and prints the difference as a .stdout difference does; --bless writes them as it writes .stdout. A .bounds or .loops.bounds with no NAME.nika is an error, as a stray .stderr is. The report is computed as check --bounds computes it for the program (the file checked as nikaia test checks it), independent of the build’s remove-* settings (ADR-306 D14: what is proved does not depend on them).
scripts/where.nika, 2026-10-31)crates/nikaia-std/src/tools/project_tests.nika: ProgramTest (50-57), program_tests (172-181, reads .stdout/.stderr with project_tests_optional), project_tests_bless_program (355-380), project_tests_difference (206), project_tests_explain_output (384). Re-lower it (lower-std) and commit lowered/project_tests.rs.crates/nikaia/src/project.rs: run_tests (about 2931) and the loop over suite.programs_of_their_own (about 3057) - the Rust side that builds and runs each program test. Edit only there: lines 2112-2586 (drive, drive_to) are cut by the main lane’s P5c move; do not touch them.crates/nikaia/src/bounds_report.rs: report (47, checks the file alone), report_checked (75); through_loops: bool is the setting.nikaia test section of docs/specification/30-nikaia-tooling.md (search .stderr near program test / ADR-269 D22): one paragraph with the decision.nikaia test: tests/compiler/tests/a_library_trait_with_no_impl_runs.nika.ProgramTest gets bounds: Bytes? and loops_bounds: Bytes?; program_tests reads them; the runner computes both reports where a file exists and fails the test on a difference, with the normalisations above; bless writes them; the stray-file error. Part III paragraph. Commit.tests/compiler/tests/bounds_expectation_guards_a_proof.nika + .stdout. In a temp project: a test program with a while counter (fn k() -> i64 sync { let mut i = 0; let mut s = 0; while i < 10 { s += 1; i += 1 }; return s }, written over lines) and .bounds/.loops.bounds that agree passes (1 passed); with proved changed to checked in one line of .loops.bounds it fails and the output names .loops.bounds; a stray x.bounds is the error. Probe first with check --bounds (with and without --optimization prove-through-loops:on) what the report really says, and write the files from that. Commit.tests/compiler/tests/prove_loop_counter_bounds.nika (p4’s probe functions k, g, f and c from docs/briefs/prover/p4-1054-loop-checks.md, Problem) and tests/compiler/tests/prove_in_loop_assert_bounds.nika (an index inside a loop under a p2 in-loop assert, e.g. while i < xs.len() { assert(i >= 0); s += xs[i]; i += 1 }), each with .stdout, .bounds and .loops.bounds written with --bless and read through: each proved there must be one the lane claims (p4’s report: k, g, f proved under on, c’s n += 1 checked). Run both by name with nikaia test --project tests/compiler --both-settings. Commit with Fixes: #1069.Changelog: this changes what nikaia test does: write the Changelog: line (one sentence), not none.
The three tests pass by name through nikaia test --project tests/compiler --both-settings, with the program tests whose names start prove_ and the tests that contain bless or stderr (by name); lower-std fixpoint clean; build, fmt, clippy (1.97 and 1.99.1 where installed), pre-push-check.sh; committed locally. Stop about 15 turns before max_turns, commit what is finished and report partial.
An edit is needed outside project_tests.nika and its lowered twin, the program-test part of project.rs, bounds_report.rs, the Part III paragraph and the new tests; an error in another agent’s file.
Not yours, never stage: every file git status shows modified or untracked at your start (crates/nikaia-std/src/tools/lowered/project_members.rs is uncommitted residue at the time of writing; do not commit it), and every docs/briefs/* file except your report.
docs/builder-facts.md first. develop in the main checkout; git status at start: other agents’ uncommitted files are not yours, never touch or stage them.git add <files> only (scripts/commit.nika -- REPORT does it); commit locally after each step and never push: the driver runs the complete suite and pushes (.claude/agents/builder.md, Cheap checks only; owner, 2026-10-31).cargo build --profile fast -p nikaia, lower-std and its fixpoint when a tools/*.nika changed, cargo fmt, cargo clippy --all-targets -- -D warnings, scripts/pre-push-check.sh, the tests the brief names and the ones your change touches. No gate.nika.CHANGELOG.md, badges; no new NK code. New tests in Nikaia only; never the full suite. No foreground sleep, no polling; scripts in files.std or the toolchain lacks stops the work: no code around it unless an ADR or Part I-III refuses it (name it). Else Result: blocked, Found: gap + probe, Needs: architect decision (CLAUDE.md, A missing form stops the work).NIKAIA_CACHE_LIMIT=2G; at most two clones with target/ in the machine; delete your own clone’s target/ and target/release/nikaia-scratch at the end; report GB freed.Result: partial with what you ruled out, the probe, and the suspected file and function (.claude/agents/builder.md).target/fast/nikaia run scripts/commit.nika -- REPORT, which writes Pushed: local <sha> into it. A report without a Pushed: line, or with a placeholder left in (TC, DISK), is not done and cuts the batch (k3a1 2026-10-29: Result: done, nothing committed)./home/user/Nikaia/docs/briefs/prover/p5-1069-bounds-expectation.md.report.md)Fields of docs/architect.md, Briefs and reports: Result: done|noop|partial|blocked (the bare word alone on that line: Result: done, never a summary after it, which cuts the batch, b869a 2026-10-27; Owner: is a question for the owner or none, never a role), Files, Commit, Body (optional, one paragraph for the commit message), Fixes, Pushed, Tests, Gate (green|red), Deviation, Found, Needs, Changelog, Owner, Disk freed, and for a move Rs2nika: (the kept N of M lines of rs2nika --score, then the three costliest hand edits; none for a task that moves no Rust, none (redesign) for a brief with Shape: redesign and no Convert: line). none where nothing, and then the bare word: Deviation: none (a note) cuts the batch (batch.nika compares exactly); a note goes into the commit message. Found: is only for a defect outside your change (it cuts the batch); a process event such as another session resetting or rewriting your files is no Found:, it goes under Needs: (m3a 2026-10-28); a converter draft that came out wrong is no Found:, it is one of the costliest edits under Rs2nika:, which the converter lane reads (e12-4b 2026-10-27); how you solved your own step goes into the commit message. Deviation: is only a step of the brief built otherwise than written or not built (a choice the brief left to you, such as keep what X needs and say so, is no deviation: it goes into the commit message; nor is building two steps in one commit when every step is built, e12-3a 2026-10-27, nor a step another commit on develop already built, shown by your test failing with that commit’s change reverted, b889 2026-10-27: name the commit in Body:, nor a test’s expected .stdout/.stderr taken from what the program really prints where the brief’s sketch of it was short or wrong, b934 2026-10-29: say so in Body:, nor a lowering that follows the ADR governing it where the brief’s expected lowering contradicts that ADR, b933 2026-10-29: name the ADR in Body:, nor a lowering other than the brief’s where rustc cannot take the brief’s, b1029 2026-10-31: say why in Body:, nor re-keying an allow-list entry (KNOWN) whose hash changed for the same program and the same reason, b-fixforward-sweep2 2026-10-28: say so in Body:, nor an existing test’s expected .stdout/.stderr updated because your change rightly changes the message it prints, b1012b 2026-10-31: name the test in Body:); how you verified (e.g. a test not run against the unfixed compiler) goes under Tests:. A noop says why under Tests: (what was probed), not under Deviation:. Final answer, one line: done <sha>, report <path> (or BLOCKED: ... / PUSH REJECTED ...).
Attribution: Co-Authored-By: Claude Opus 5.5 noreply@anthropic.com Claude-Session: https://claude.ai/code/session_01UVxJDtwVhc6jMtEYAyvwLU