Ship 3.5c's provenance attribution to the central learner. ExportLabel gains two
additive uint64 tags (8 origin_node_hash, 9 origin_exe_hash); a positive window's
label now carries token_hash(dropper node) + token_hash(path_template(dropper
exe)) from ReplayWindow.origin_*. fleet_merge counts windows_with_origin.
R12/R13: a FIELD addition, not a new record type -- no codec.hpp/enum change
(R12 untouched), both fields are uint64 hashes so no raw path/node string crosses
the export boundary (R13 green, no name-loophole). Back-compat additive:
kExportSchemaVersion stays 1; old bundles decode with origins 0; old readers
ignore tags 8/9. Positive labels only -- suppressed/negative carry no origin.
Tests: round-trip of the hash fields + defaults; extraction proves a dropper->exec
detected window's positive label carries a non-zero origin while a no-write
detected window and suppressed labels carry 0. source_invariants green.
A learned ranking head + TrainingExample plumbing are phase-2 (out of scope).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reuse the global write->exec linkage (recent_writes) to attribute a detection to
its ROOT CAUSE. recent_writes now maps path -> {writer_node, writer_exe}; when a
window is late_detected (retro-hunt flags a binary after it ran) or detected
(YARA), set ReplayWindow.origin_node/origin_exe from whoever WROTE the flagged
binary -- usually an earlier node than the exec leaf that ran it. Surfaced in
window_json. Deterministic provenance walk (first writer wins); replay-level only,
no export-schema/record-type/enum change (R12/R13 intact), self_written_execs
unchanged. A learned ranking head over ambiguous origins is the follow-up (3.5d).
Test R10: a dropper writes /tmp/evil in one window, a separate process execs it
and is flagged late -> the window's origin_node is the dropper, not the exec leaf;
no in-set write -> empty origin; deterministic.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
#4 (guard retention): same-binary parent/child self-ptrace never rates "high"
(it's self-debugging, not injection), but from a drop location (/tmp) it now
stays "medium" rather than "low" -- a same-binary anti-debug/injection staged in
/tmp still surfaces, while sanitizer builds no longer flood high-severity
PtraceAttach. Injection into another program is unchanged.
#5 (CLI-UX contract): --max-node-share/--test-fraction/--split now error when
passed to any command other than 'fleet train' instead of being silently
ignored (noun-verb commands already reject unknown options; this covers the
flag-form and other fleet subcommands).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Address code-review findings on this session's changes (no behavior change):
- Share the dense-slot index for the write->exec feature as
learn::kWriteExecFeatureIndex; featurize() and the model-score dev tool now
reference it instead of a duplicated literal 42 (prevents silent column drift
if the feature layout shifts).
- Fix the self_written_execs comments: the linkage is a GLOBAL bounded FIFO
(recent_writes), not strictly "this window" — the docs now match replay.cpp.
- release.yml: never cancel a tag/release run (cancel-in-progress only for
non-tag refs); push ordering doesn't guarantee the tag event arrives after the
branch event, so a same-sha branch push must not abort an in-progress release.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The version lived in three hard-coded copies (CLI `version`, daemon status,
banner); v0.2.0 shipped with the CLI still saying 0.1.0. meson.build is now
the only source: it passes STRAWBERRY_VERSION to every target, and the CLI
test requires the exact version.
packaging/ci/build-release.sh refuses to publish a tag build unless
meson.build, the tag and the built binary's `strawberry version` agree; CI
never rewrites the version.
scripts/release.sh X.Y.Z is the one manual release step: on a clean main
equal to forgejo/main it bumps meson.build, makes a signed commit and a
signed tag, and pushes both (--dry-run shows the plan). The Security Hub
release channel deploys signed tags only.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A parent/child of the SAME binary tracing each other is self-debugging
(sanitizers' StopTheWorld, crash handlers), not injection. The code already
scored that `low`, but `suspicious_image` (which flags drop locations like /tmp)
took precedence and forced `high` -- so sanitizer test binaries built in /tmp
flooded findings as high-severity PtraceAttach (observed live: ~160 FPs on
/tmp/toolcheck-*/*_san). Reorder so the strong structural self-debug signal beats
the weak location heuristic: same-program-family -> low regardless of location;
ptrace into ANOTHER program stays high (suspicious image) or medium. Injection
detection is unchanged. Test added for the /tmp self-ptrace case.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Build-only tool: loads a trained model (.sbm) and an export bundle, applies
forward(model, features) per window, prints TSV window_id/value/label/x[42] plus
a summary partitioned by the write->exec feature slot. Unblocks the novel-chain
benchmark: replay --model ranks by rarity (model_value null), so it could not
measure a model's score on holdout windows; this does. On the detector-covered
train bundle the data-flow model scores write->exec windows 0.634 vs 0.0003 for
the rest -- it clearly learned the x[42] feature.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
3.5b's ordered exec-lineage proved inert in clean A/Bs; replace it with a general
write->exec data-flow feature: WindowSummary.self_written_execs (exec of a file
written earlier in the same window) at x[42], tracked via OpenWindow.written_paths.
Honest status: this is the conceptually-correct provenance feature (it separates
download->exec-the-same-file from download->exec-unrelated), but it ALSO shows no
AUC gain on detector-labeled data -- for the same structural reason as 3.5b: the
teacher labels (DownloadThenExecute etc.) correlate with existing count/location
features or re-derive the detector. Validating provenance LEARNING requires the
novel-chain generalization benchmark (train on detector-covered, hold out
detector-evading attacks); committed as that benchmark's feature foundation.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Feed ordered parent->child exe-template bigrams into the window feature vector
(WindowSummary.lineage_edges, note_lineage_edge, a salted hash loop in featurize,
wired at replay on_exec). The model already got per-kind lineage surprisal
(x[25..29]); this adds path ORDER, which the unordered token bag discarded.
Clean A/B on a fixed corpus (same 697 windows, same flags, only the binary
differs): model AUC 0.734 -> 0.737 (rarity 0.520 unchanged). Null on this
shallow-lineage data (marker execs are 1-process, synthetic data flat) but the
feature is correct and order-sensitive (test: A>B != B>A). Kept as the path
representation that Stage 3.5c's retro-hunt attribution head needs; its payoff is
expected on real multi-stage attack telemetry.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Expose the merge cap and train/test split on `strawberry fleet train` so a
single-node real-attack bundle can pre-seed a model. Default stays temporal +
cap 0.25 (behavior-preserving); `--split random` + `--max-node-share 1` lets the
Caldera-attack export (15 positive windows) train. Verified: real-only train AUC
1.0 vs rarity 0.563; default still refuses identically.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Do not turn an empty cached digest into an unbounded descriptor hash. Distinguish size refusals from YARA results so the exec guard applies its own bounded scan/degrade policy. Cover cached read budgets across size boundaries, known hashes, inventory/live events and real guard-cache reuse. Fold/live/KPI and post-rollout CPU acceptance remain pending; OpenSnitch checksum work is separate.
Use suspend-aware BOOTTIME waits and owned-container cleanup without changing the workload. Reject missing coverage and preserve producer/report failures. Include runtime/helper regressions, constrained build-job support and live exec-guard crash safety. Document the failed live/idle gates without relaxing thresholds or claiming production acceptance.
Preserve independent support, typed rule namespaces and stable process lifetimes. Share authoritative managed-set reconciliation across daemon and offline correlation, load offline policy without seeding/migration, and release resources after shutdown-flush exceptions. Keep retirement history, transaction ownership and fault/allocation regressions; no whole-set crash-atomicity or live acceptance claim.
Reuse the existing resolver/sampler for opt-in aggregate training summaries. Reject incomplete or undecodable snapshots without usable partial reports, retain best-effort readers elsewhere, and document cooperative extent checks rather than atomic snapshot guarantees. Include command-table/help, JSON, fixture and corruption regressions; no model training or promotion.
Represent full-range ordered clock differences without signed overflow and allow complete decay. Add seeded restore-clock regressions and a dedicated fuzz seed.
Allowlisting a binary suppressed only *new* detections: the findings
already stored for it stayed in the evidence store, and every reload
re-materialised them. That is why the ~9k iproute2 `ss` false positives
kept "coming back" after the hash was allowlisted a day earlier -- nothing
retracted them, and the only cleanup was `doctor fix
retire-allowlisted-findings`, which operators do not know to run.
allowlist-add now retires the now-allowlisted findings at the moment of
the action (alongside the existing containment release) and reports the
count as `retired_findings` in its reply. retire_allowlisted_findings is
hash-based and self-filtering, so a path/rule entry that matches no
fully-allowlisted finding retires nothing.
Regression test (test_daemon_runtime): a detected binary's finding is
materialised (forced correlation), allowlist-add reports
retired_findings >= 1, and after correlation re-runs the finding does not
come back (A already makes correlation skip allowlisted entities).
Also refresh the doctor findings.allowlisted message, which claimed
allowlisting does not delete stored findings -- it now does.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On a host that sits at critical/emergency for a long time (its own apps holding
most of RAM, or the daemon slightly over its soft budget), the starvation guard
was disabled (`!hard`), so correlation, retro-hunt and inventory deferred
forever — observed deferred 153 min with forced=0, which also starves the
retro-hunt delayed labels that Stage 4 training needs.
- Tiered starvation guard: it now fires under critical/emergency too, but after a
level-scaled window (1x warning, 4x critical, 12x emergency) so background work
is rare under load, not zero. A true host-memory crunch (host_severe_,
MemAvailable below the severe floor) still holds everything, since forcing work
then could tip the host toward OOM. exec guard/scan paths unchanged.
- Raise the balanced own-memory soft budget 512 -> 768 MiB: the daemon's steady
footprint grew (telemetry buffers, the ingest server, the shadow model), so 512
left it permanently over budget -> pinned critical/emergency. 768 fits the real
baseline, still well under the 1.5 GiB cgroup cap.
Tests updated for the 768 budget and the tiered guard (critical forces after the
4x window; severe holds it). Suite 51/51; analyzer + clang-tidy clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The teacher-label join made a window a positive if it was detected
(replay.cpp), but it never applied the allowlist, so a window running a cleared
false positive (e.g. /usr/bin/ss) was labelled a teacher positive — poisoning
Stage 4 training. Negative labels were written to the stream but replay dropped
them.
- replay_store now collects negative labels into ReplayOptions.negative_labels
(it used to ignore any non-positive label).
- A window whose exec hash is negatively labelled is marked `suppressed`; the
example/policy-eval positive is gated on `!suppressed`, so such a window is
kept as a *negative* training example (useful signal) and never a positive,
even if a cleared rule "detected" it.
- DetectionPipeline seeds a negative label for every allowlisted sha256 at init,
so evidence stored before the allowlist (the historical ss storm) is covered
without waiting for a re-exec.
Test: a detected window with a stored negative label for its hash is a positive
before the label and a kept-negative after. Suite 51/51; analyzer + clang-tidy
clean on replay.cpp and detection_pipeline.cpp.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Correlation runs over the whole store and never consulted the allowlist, so an
allowlisted hash (e.g. the /usr/bin/ss feed false positive the operator cleared)
regenerated the same findings on every pass from already-stored indicators —
`doctor fix retire-allowlisted-findings` cleared them but they came straight back
(~9,400 findings observed again). This also meant the findings were a poisoned
teacher-label source for training.
- classify() now skips an indicator whose (hash, path, rule) is allowlisted, the
single chokepoint shared by CorrelationEngine::build_findings (the oracle) and
IncrementalCorrelator::update, so both agree. New AllowlistFn; build_findings /
correlate take an optional predicate (default = no filtering, existing tests and
the incremental==full oracle unchanged); IncrementalCorrelator::set_allowlist.
- DetectionPipeline wires the predicate to PolicyStore::match, so the live
incremental path skips allowlisted evidence. Retiring the findings now sticks.
Test: build_findings with an allowlist predicate emits nothing for the matched
hash and is unaffected for others. Suite 51/51; analyzer + clang-tidy clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Completes the tracker-attribution transport on macOS. There is no network
sandbox to work around there, so the daemon (root) reads Little Snitch's traffic
stream directly and feeds the same on_opensnitch_connection consumer — no relay
or ingest socket (those are the Linux sandbox answer).
- LittleSnitchParser (cross-platform, unit-tested): parses `littlesnitch
log-traffic` CSV via a header-name->index map (tolerates column reordering),
yielding the shared redacted OpenSnitchConnection for outbound rows only
(program, uid, host/ip/port, protocol 6->tcp/17->udp). Little Snitch's CSV has
no argv/env, so redaction is structural. split_csv_line handles quoted fields.
- LittleSnitchSource (macOS only): posix_spawns `littlesnitch log-traffic
--stream`, reads the pipe on a guarded thread, parses and calls back; stop()
SIGTERMs the CLI (EOF -> loop ends) then reaps it.
- Daemon: start_tracker_attribution gains an __APPLE__ branch that starts the
source; shutdown/status handle it. New config littlesnitch_cli (default CLI
path). status privacy.listener source "little-snitch".
- docs: macOS section rewritten (needs Little Snitch "Allow access via Terminal").
Test E10 covers the parser (header, outbound/inbound, deny, quoted exe, no-dest).
Linux suite 51/51; analyzer + clang-tidy clean (Linux view). The __APPLE__ code
compiles/validates on the Mac next (ssh); live spawn needs root + LS Terminal access.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
First half of the sandbox-safe tracker transport (architecture deep dive). The
daemon's old loopback TCP listener could not bind under its own sandbox
(PrivateNetwork=true + RestrictAddressFamilies=AF_UNIX). Replace it with an
AF_UNIX ingest socket the daemon *can* bind under the sandbox; a host-namespace
relay (next change) will bind the OpenSnitch port, strip argv/env, and forward
redacted tuples here.
- opensnitch_events.hpp/cpp: encode_connection / decode_connection serialize only
the redacted fields (never argv/env, which the parser already drops) as one
JSON object — the relay->daemon wire form. New TrackerIngestServer: an AF_UNIX
stream listener (Linux-only, like the TCP listener) that accepts only a root
peer (SO_PEERCRED), frames with take_json_objects, decodes and calls back;
bounded clients/object/buffer as before.
- daemon: start_tracker_attribution now opens the ingest socket
(opensnitch_ingest_socket, default /run/strawberry/trackers.sock) instead of
the in-sandbox TCP listener, so it works with the sandbox unchanged and is
simply idle until a relay connects. status privacy.listener reports the ingest
socket + relay_events; shutdown stops it. New config key opensnitch_ingest_socket.
- Tests: E8 encode/decode round-trip + redaction (no args/env on the wire);
E9 TrackerIngestServer end to end (connect, two tuples, callback, stats);
daemon-runtime configs set privacy_trackers=false (not under test).
Suite 51/51; clang static analyzer and clang-tidy clean on the changed TUs. The
relay run-mode + systemd unit + install wiring + docs follow in WS2b/WS2c.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The scan path (handle_process) and the exec-guard recording path
(record_exec_decision) had grown two near-identical repeat-aggregation maps
(repeats_ / exec_repeats_) plus three copies of the IOC-feed severity idiom.
Refactor to one implementation; no behavior change (the existing aggregation and
IOC-severity tests are the oracle and pass unchanged).
- New include/strawberry/repeat_aggregator.hpp: RepeatAggregator owns the
per-key window map, the 20-full-then-summary cadence and the aggregated count;
observe(key, now, seen_ns, pid) returns {full, at_threshold, aggregated,
emit_summary, notify, state}. Each caller keeps its own key, summary-id prefix,
summary shape and notify behaviour. Both pipeline sinks now hold one instance
(scan notifies at most hourly; exec passes notify_ns=0). status
repeats_aggregated sums both totals (atomic, lock-free read).
- ioc_hit_severity(feed): one source for the three IOC match sites (scan_image,
assess_exec, handle_process) that was the policy_->ioc_feed_severity ternary.
- Leaves the latency-critical assess_exec/decide path untouched; both sinks still
run under mutex_, off the FAN response.
Unit test tests/test_repeat_aggregator.cpp (window reset, threshold, summary and
notify cadence, per-key independence). Suite 51/51; clang static analyzer and
clang-tidy clean on detection_pipeline.cpp; fold's 7 int-overflow are the accepted
monotonic-timestamp FP class (unchanged by the move); perf-kpi PASS, no regression
(exec_decision p50 unchanged, exec_assess/db_commit within noise or faster).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Allowlisting a hash suppresses new detections but does not delete findings
already stored for it (e.g. the /usr/bin/ss false-positive storm). That cleanup
now lives in doctor instead of a one-off compaction.
- DetectionPipeline::retire_allowlisted_findings(dry_run): a finding is stale
when it references at least one hash and every hash it references (from
metadata "hashes" and file_<sha256> nodes) is allowlisted, so a finding that
also implicates a non-allowlisted binary is never dropped. Mirrors
retire_superseded_findings (retire_finding + transaction).
- node_ids.hpp gains content_hash_of_file_node_id() (the inverse of
file_node_id for the content case) so the pipeline parses file ids through the
canonical helper, not a "file_" literal (source invariant R8).
- Doctor check findings.allowlisted (warn) with maintenance action
retire-allowlisted-findings; DoctorInputs.allowlisted_findings gathered by the
daemon; run_maintenance handles the action; CLI `doctor fix` help + completion
and API_CONTRACT updated.
Tests: test_retire_allowlisted_findings (two execs of a flagged binary -> a
finding; allowlist the hash -> dry-run lists it, apply retires it, idempotent);
doctor check ok/warn cases; CLI completion string. Suite 50/50; clang static
analyzer, clang-tidy and source invariants clean on the changed TUs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The exec-guard would-block recording path stored one indicator (and, via
correlation's per-process grouping, one finding) per exec, with no aggregation,
unlike the scan path. A critical feed or rule matching a frequently-run binary
would flood evidence the way /usr/bin/ss did.
- record_exec_decision now collapses repeated identical decisions (keyed by
verdict|hash|rule) with the same 10-min window / 20-full-then-summary scheme
as handle_process (new exec_repeats_ map; counts into repeats_aggregated_).
Beyond the window it refreshes one "exec-repeat" summary indicator and returns
before the per-exec graph write.
- The actor's /proc identity reads are now done before taking mutex_, not under
it. The clang static analyzer (new strict gate) flagged ::read in a critical
section (unix.BlockInCriticalSection) here; this also honors the AGENTS rule
that scans/reads stay off DetectionPipeline::mutex_. Analyzer now clean.
Test: test_repeated_would_block_is_summarized drives 30 identical would-block
decisions and asserts exactly 20 full exec-audit indicators + 1 exec-repeat
summary and repeats_aggregated == 10. Suite 50/50; analyzer and clang-tidy clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A benign system binary (/usr/bin/ss) whose SHA-256 was on a noisy community hash
feed produced thousands of critical "would have been blocked" findings: IOC hash
matches were hardcoded severity "critical" at every site, and in exec_guard an
ioc_match forces a prevention hit regardless of the severity threshold.
- FeedSpec gains `max_severity` (append-only; serialize/parse/validate, control
socket `feed-add` accepts it). Empty = default for the kind: hash feeds are
advisory "high", not "critical". PolicyStore::ioc_feed_severity() resolves it.
- All three IOC severity sites now use the feed's severity
(detection_pipeline.cpp scan_image, assess_exec, handle_process), replacing the
hardcoded "critical".
- Trust gating: ExecAssessment::ioc_match is set only when the matched feed
resolves to "critical" (a curated feed that should force prevention). A
community feed stays advisory — its "high" is subject to prevention_min_severity
(default "critical"), so it no longer would-blocks system tools per exec.
Tests: test_ioc_hash_detection uses an explicit critical feed (trusted path) and
asserts assess_exec forces the hit; new test_ioc_community_feed_is_advisory
asserts a default feed is "high", ioc_match is not forced, and decide_exec allows
it in audit and enforce; test_policy covers max_severity validation, the advisory
default, and a critical round-trip. Suite 50/50; clang static analyzer and
clang-tidy clean on the changed TUs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Second piece of tracker attribution (docs/tracker-attribution.md): which
programs reach domains on a trackers list. Advisory only: never a block,
never a finding (source invariant R11).
- OpenSnitch event listener: OpenSnitch's JSON "remote" logger is the
hostname source (Strawberry's eBPF connects have none). It streams over
TCP and cannot leave fields out, so events carry command lines and
environments. The listener binds 127.0.0.1 at a port below 1024 (only
root can), accepts only root-owned peers (/proc/net/tcp), bounds
clients/object size/buffering, and reads only pid, uid, path,
destination and action; arguments and environment are never copied.
- TrackerContacts: per program x domain, bounded at 20000 keys. A root-only
tracker_contact telemetry record (type 25) on the first contact, then at
most hourly with the count since.
- Status `privacy` block; control verb privacy-trackers; CLI
`strawberry privacy trackers [--json]`.
- `strawberry opensnitch enable-events` adds the logger to OpenSnitch's
config (with a .bak); the installer runs it when OpenSnitch is present,
unless privacy_trackers=false. Config: privacy_trackers (default true),
opensnitch_events (default 127.0.0.1:979).
- EasyPrivacy is a default trackers feed (GPL-3.0+ or CC-BY-SA-3.0,
downloaded at run time).
Tests: E1-E7 (test_opensnitch_events: event fields, stream splitting,
peer uid, address rules, listener end to end, contacts and record, config
editor), the control verb, CLI translation and help, daemon status. New
fuzz_opensnitch_events (90 s clean).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
First piece of tracker attribution (docs/tracker-attribution.md): which
programs reach advertising and tracking domains, built with the
measurement rules of the CarolinaCon 2026 privacy talk (bridgescope):
a listed domain is a circumstantial label, never proof.
- Feed kind "trackers": hosts lines, plain domains and adblock
whole-domain rules (||host^, ||host^$third-party). Cosmetic rules,
exceptions, paths and other options are skipped; the comment cut used
to turn "site.example##.ad" into a domain.
- TrackerSet: suffix matching (a.b.t.example matches t.example, never
nott.example, an IP literal or a bare TLD); first feed reported.
- List parsing moves from rule_updater.cpp to ioc_list.cpp (no network
code) so the query-side core can use it.
- Source invariant R11: tracker lists never meet exec-guard or OpenSnitch
code. They are not in the OpenSnitch blocklist sync.
Tests: T1-T3 (test_tracker_set), trackers in the list-parsing property
and the list fuzzer (60 s clean), R11 self-test fixture.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
`model train` on a host without detections stops with "0 positive
windows" and pointed at --inject-train FILE, but only the evaluation set
(standard.trace) existed. Add config/traces/training.trace: 7 attack
shapes (document macro dropper, web reverse shell, cron miner, /dev/shm
perl stager, ssh key backdoor, npm postinstall, ld.so.preload) that differ
from the evaluation set, so the gate never judges a model on what it
trained on. P4 in test_trace enforces that (no shared trace name or exec
sequence); it caught an evince -> python3 overlap in the first draft.
Both installers put the traces in /usr/local/share/strawberry/traces/
(install_traces in install_common.sh); the hint and README name it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The shadow scorer's rarity model started empty at every daemon start, so
each restart cost an hour of warm-up with no decisions. The daemon now
saves it to model_dir/rarity.state (checksummed, atomic write, hourly and
at stop) and restores it at start. A restored state younger than the
7-day half-life skips the warm-up; decay catches up on the first event.
A file made with other rarity settings, a bad checksum, or a negative or
non-finite count is refused with a reason (status model.rarity_state,
doctor model.rarity_state) and the model starts empty. The file must be
owned by the daemon user and not group/world writable, like model files.
Also fixes status model.warmup_windows, which was never filled in.
Tests: R1/R2 in test_learn (round trip, decay clock, refusals), S8 in
test_shadow (warm-up skipped only when recent), daemon restart restores
the saved state, doctor warning; new fuzz_rarity_state target.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>