pi-lens 4.3.0 → 4.4.1

package/dist/clients/cache-observability.js

@@ -1,1313 +0,0 @@
-/**
- * Prompt-cache observability (#1018).
- *
- * Two provider-independent signals, both sinking into ~/.pi-lens/latency.log
- * via {@link logLatency} (matching the neighboring `type: "phase"` records):
- *
- *   1. Response-side cache usage (`cache_usage`) — on each assistant
- *      `message_end`, record the provider-reported token/cost breakdown so a
- *      session's actual cacheRead/cacheWrite behavior is queryable after the
- *      fact. This is what the provider DID, not what we hoped it would do.
- *
- *   2. Request-side context observations (`cache_context`) — one bounded record
- *      for every `context` call, describing pi-lens's own before/after message
- *      transformation, placement, injection sources, and privacy-preserving
- *      hashes. This is what pi-lens observed locally, not the final provider
- *      request after other context handlers have run and not a provider cache
- *      result. #1938: this record used to also carry a `prefixObservation` /
- *      `firstMessageChange` pair derived from a SEPARATE bounded hash of the
- *      pre-injection prefix (capped at `MAX_HASHED_MESSAGES` messages /
- *      `MAX_HASHED_CONTENT_CHARS` chars). Every real session's transcript
- *      exceeds that cap almost immediately, so the pair reported "unknown" on
- *      97% of records — dead weight in every record, answering nothing. The
- *      pair is removed; signal 3 below (`cache_prefix_break`) already covers
- *      first-message stability with an unbounded hash and never truncates.
- *
- *   3. Request-side prefix stability (`cache_prefix_break`) — a content hash of
- *      `messages[0]` observed on every `context` call. After #1016 the first
- *      message must stay byte-stable across a whole session; a logged CHANGE
- *      flags that something (pi-lens or otherwise) broke the cache prefix. This
- *      remains a local observation, never a provider cache-miss claim.
- *
- * #1071 adds miss ATTRIBUTION on top of those three. Each `cache_usage` record
- * carries the inter-turn gap and a `cacheMissCause` verdict, and each
- * `cache_context` record splits a mixed injection payload by contributing
- * source. #1996 makes every unexplained miss name the absent/conflicting
- * evidence and emits one fixed-key `cache_usage_summary` at session shutdown.
- * Heuristic causes require complete request hashing and full provider/model
- * identity evidence. Identity comparison uses SHA-256 over the full normalized
- * value while logs retain only a bounded display prefix; malformed numeric
- * usage is sanitized before it reaches JSON metadata.
- * The per-turn records still ride events that already fire once per turn, and
- * all attribution derives from bounded in-process state — nothing reads
- * latency.log back or serializes transcript content.
- *
- * All paths are defensive: `usage` (and its fields) may be absent on older
- * hosts or non-assistant messages, so every access is guarded and the handlers
- * never throw — on error they `dbg(...)` and no-op, like the other index.ts
- * event handlers.
- */
-import { createHash } from "node:crypto";
-import { BoundedFifoMap, BoundedLruCache } from "./bounded-cache.js";
-import { lazyEnvNumber } from "./env-utils.js";
-import { logLatency } from "./latency-logger.js";
-import { compareOrdinal } from "./string-utils.js";
-/** RuntimeCoordinator's counter is process-global when sessions share a host. */
-const PROCESS_TURN_SCOPE = "process-global-runtime";
-const SECONDARY_TURN_SCOPE = "unavailable-concurrent-secondary";
-/**
- * Keep context telemetry bounded even when a tool result or user message is
- * unexpectedly large. This is a hash input only; none of the sampled text is
- * written to the log.
- */
-const MAX_HASHED_MESSAGES = 64;
-const MAX_HASHED_CONTENT_CHARS = 2048;
-const MAX_REPORTED_MESSAGES = 10_000;
-const MAX_INJECTED_CHARS = 16_384;
-const MAX_INJECTED_BYTES = 65_536;
-let contextObservationCounter = 0;
-/**
- * Documented deterministic char→token estimate. Four characters per token is
- * the same ratio #1071's issue body used to size the sampled injections. It is
- * an ESTIMATE: never present a value derived from it as provider-measured
- * usage.
- */
-const CHARS_PER_ESTIMATED_TOKEN = 4;
-/**
- * Idle gap after which a provider prompt cache entry is assumed expired.
- *
- * The gap this is compared against is measured to REQUEST time, not to the
- * `message_end` that follows it — see {@link resolveInterTurnGap}. The provider
- * looks the cache up when the request arrives, so the response's own generation
- * time is not idle time and must not count toward expiry.
- *
- * Evidence (2026-08-21 KV-cache investigation, 63 sessions, 2,288 turns): the
- * 34 zero-`cacheRead` turns had a MEDIAN inter-turn gap of 166s, against 9s for
- * every other turn. Those 34 turns carried 6.96M fresh input tokens, 35.1% of
- * all fresh input in the window. 60s sits between the two populations and
- * matches the documented 5-minute-minus-safety-margin behavior of the shortest
- * provider TTL tier we run against. Override with
- * `PI_LENS_PROVIDER_CACHE_TTL_MS` when a provider's TTL differs.
- */
-export const DEFAULT_PROVIDER_CACHE_TTL_MS = 60_000;
-const _providerCacheTtl = lazyEnvNumber("PI_LENS_PROVIDER_CACHE_TTL_MS", DEFAULT_PROVIDER_CACHE_TTL_MS);
-const getProviderCacheTtlMs = _providerCacheTtl.get;
-export const _resetProviderCacheTtlForTests = _providerCacheTtl._resetForTests;
-/**
- * A turn counts as "anomalously low" when it reads back less than half of what
- * the previous turn read. The investigation's partial-break population (90
- * turns, upper-bounded at 2.7M tokens) was selected by exactly this shape: a
- * non-zero `cacheRead` well below the prior turn's, which a zero-read filter
- * misses entirely.
- */
-const LOW_CACHE_READ_RATIO = 0.5;
-/**
- * Fresh input must exceed the estimated new content by this factor before a
- * shortfall reads as provider-side eviction rather than as our own payload.
- * The investigation upper-bounds a whole session's injections at ~2.7 tokens
- * per turn, so fresh input several times the measured new content cannot be
- * explained by what pi-lens added.
- */
-const PARTIAL_EVICTION_INPUT_FACTOR = 4;
-/**
- * Floor for the same rule, so a turn with almost no new content does not read
- * as eviction on a handful of tokens.
- */
-const PARTIAL_EVICTION_MIN_FRESH_TOKENS = 512;
-/**
- * Cap on the per-session character accumulators used for attribution, and on a
- * single transcript-growth measurement feeding them.
- *
- * This is deliberately NOT `MAX_INJECTED_CHARS`. That 16 KiB cap exists to keep
- * the reported injection size small, and pi-lens injections average about 2.7
- * estimated tokens per turn, so it is never reached in practice. Transcript
- * growth is a different population: one ordinary tool result can add 20,000
- * characters in a single turn. Measuring growth through the injection cap
- * latched `attributionCharsCapped` on a routine turn and suppressed the
- * `partial-eviction` verdict, which made the verdict near-unreachable (#1071
- * review round 1, F2).
- */
-const MAX_ATTRIBUTION_CHARS = 262_144;
-/** Byte companion to {@link MAX_ATTRIBUTION_CHARS}, at four bytes per char. */
-const MAX_ATTRIBUTION_BYTES = 1_048_576;
-function emptyCacheMissCauses() {
-    return {
-        "ttl-expired": 0,
-        "prefix-broke": 0,
-        "partial-eviction": 0,
-        "model-provider-changed": 0,
-        unknown: 0,
-    };
-}
-function emptyUnknownEvidenceReasons() {
-    return {
-        "no-prior-sample": 0,
-        "cache-read-unavailable": 0,
-        "malformed-provider-usage": 0,
-        "request-correlation-unavailable": 0,
-        "request-evidence-incomplete": 0,
-        "sequence-hash-truncated": 0,
-        "model-provider-unavailable": 0,
-        "provider-reported-zero-stable-prefix": 0,
-        "provider-reported-low-read-stable-prefix": 0,
-        "no-local-explanation": 0,
-    };
-}
-function newCacheUsageSummary() {
-    return {
-        usageRecords: 0,
-        cacheHits: 0,
-        missObservations: 0,
-        cacheMissCauses: emptyCacheMissCauses(),
-        unknownEvidenceReasons: emptyUnknownEvidenceReasons(),
-    };
-}
-// Shared cap for every per-session bounded map in this module (also used by
-// `prefixHashBySession` below) — a long-lived process cycling through many
-// sessions cannot grow any of them without limit.
-export const MAX_TRACKED_SESSIONS = 32;
-const attributionBySession = new BoundedLruCache(MAX_TRACKED_SESSIONS);
-function newAttributionState() {
-    return {
-        prefixBrokeSinceLastUsage: false,
-        prefixStableSinceLastUsage: false,
-        sequenceHashTruncatedSinceLastUsage: false,
-        requestEvidenceIncompleteSinceLastUsage: false,
-        injectedCharsSinceLastUsage: 0,
-        newTranscriptCharsSinceLastUsage: 0,
-        attributionCharsCapped: false,
-        injectedBytes: emptyInjectedBytes(),
-        injectedFindingsRepeated: 0,
-        findingIdentities: new Set(),
-        findingIdentityCapRecorded: false,
-        turnEndAdvisoryBytes: 0,
-        toolResultBytes: 0,
-        toolResultsTruncated: 0,
-        summary: newCacheUsageSummary(),
-    };
-}
-const MAX_FINDING_IDENTITIES = 2_000;
-function emptyInjectedBytes() {
-    return {
-        sessionGuidance: 0,
-        turnFindings: 0,
-        testFindings: 0,
-        agentNudge: 0,
-        turnEndAdvisory: 0,
-        other: 0,
-    };
-}
-function sourceKey(source) {
-    return {
-        "session-guidance": "sessionGuidance",
-        "turn-findings": "turnFindings",
-        "test-findings": "testFindings",
-        "agent-nudge": "agentNudge",
-    }[source];
-}
-/** Best-effort, language-neutral identity extraction from rendered findings. */
-function findingIdentities(text) {
-    const out = [];
-    for (const line of text.split(/\r?\n/)) {
-        const location = line.match(/(?:^|\s)([^\s:]+):(\d+)(?::\d+)?/);
-        if (!location)
-            continue;
-        const rule = line.match(/(?:rule|code)\s*[:=]\s*([\w./-]+)/i)?.[1] ??
-            line.match(/\(([\w./-]+)\)\s*$/)?.[1] ??
-            line.match(/:\d+(?::\d+)?\s+([\w./-]+)/)?.[1];
-        if (rule)
-            out.push(`${location[1]}:${location[2]}:${rule}`);
-    }
-    return out;
-}
-/**
- * Fetch (creating if absent) one session's attribution state, refreshing LRU
- * recency and evicting the oldest entries past the cap — same bounding as
- * {@link recordSessionHash}, so a long-lived process cycling through sessions
- * cannot grow this map without limit.
- */
-function attributionFor(key) {
-    const existing = attributionBySession.get(key);
-    const state = existing ?? newAttributionState();
-    // `get` above already moved an existing entry to MRU; `set` below also
-    // refreshes recency on a create, so every fetch counts as a touch.
-    attributionBySession.set(key, state);
-    return state;
-}
-/**
- * Fold one `context` observation into the session's attribution state: how much
- * pi-lens injected, and how much the transcript itself grew since the previous
- * observation. The transcript delta is what the "far above the injected and new
- * content" rule measures fresh input against. The first observation in a
- * session only anchors the message-count baseline, because there is no earlier
- * point to measure growth from.
- */
-function recordContextAttribution(key, existingMessages, injectedSize) {
-    const state = attributionFor(key);
-    const baseline = Math.min(state.lastObservedMessageCount ?? existingMessages.length, existingMessages.length);
-    const grown = measureTranscriptGrowth(existingMessages.slice(baseline));
-    state.injectedCharsSinceLastUsage = Math.min(MAX_ATTRIBUTION_CHARS, state.injectedCharsSinceLastUsage + injectedSize.chars);
-    state.newTranscriptCharsSinceLastUsage = Math.min(MAX_ATTRIBUTION_CHARS, state.newTranscriptCharsSinceLastUsage + grown.chars);
-    if (injectedSize.capped ||
-        grown.capped ||
-        state.injectedCharsSinceLastUsage >= MAX_ATTRIBUTION_CHARS ||
-        state.newTranscriptCharsSinceLastUsage >= MAX_ATTRIBUTION_CHARS) {
-        // The accumulators are now floors, not totals. The verdict rules must not
-        // treat a floor as a measured value.
-        state.attributionCharsCapped = true;
-    }
-    state.lastObservedMessageCount = existingMessages.length;
-    // #1071 review round 1, F1: this handler runs at REQUEST time, which is when
-    // the provider looks the prompt cache up. Stamping it here is what lets the
-    // gap exclude the response's own generation time.
-    state.lastContextAtMs = Date.now();
-}
-/**
- * Idle milliseconds before this turn's provider request.
- *
- * Measuring `message_end` to `message_end` would fold the response's own
- * generation time into the gap, and generation dominates: over a real corpus
- * the median generation time is 7,407 ms against a median true idle of 228 ms,
- * a factor of 30. At the 60s threshold, 8.7% of consecutive pairs measure over
- * on the message-end basis while only 3.3% were truly idle over. That is a 3.4x
- * inflation of `ttl-expired`, which would have confirmed the investigation's
- * hypothesis by construction (#1071 review round 1, F1).
- */
-function resolveInterTurnGap(state, nowMs) {
-    if (state.lastUsageAtMs === undefined) {
-        return { gapMs: null, basis: "no-prior-turn" };
-    }
-    if (state.lastContextAtMs !== undefined) {
-        return {
-            gapMs: Math.max(0, state.lastContextAtMs - state.lastUsageAtMs),
-            basis: "request-time",
-        };
-    }
-    return {
-        gapMs: Math.max(0, nowMs - state.lastUsageAtMs),
-        basis: "message-end-fallback",
-    };
-}
-function estimateTokens(chars) {
-    return Math.ceil(chars / CHARS_PER_ESTIMATED_TOKEN);
-}
-function readableTokenCount(value) {
-    return typeof value === "number" && Number.isFinite(value) && value >= 0
-        ? value
-        : undefined;
-}
-function readNumericEvidence(record, key) {
-    let raw;
-    try {
-        raw = record[key];
-    }
-    catch {
-        return { status: "malformed" };
-    }
-    if (raw === undefined)
-        return { status: "absent" };
-    const value = readableTokenCount(raw);
-    return value === undefined
-        ? { status: "malformed" }
-        : { status: "available", value };
-}
-const MAX_LOGGED_IDENTITY_CHARS = 200;
-/**
- * Keep the human-readable identity bounded while comparing a SHA-256 digest of
- * the FULL normalized value. Comparing the display prefix would make two long
- * provider/model ids that differ after the cap collide (#1996 review).
- */
-function identityEvidence(value) {
-    if (typeof value !== "string")
-        return { status: "unavailable" };
-    const normalized = value.trim();
-    if (!normalized)
-        return { status: "unavailable" };
-    const truncated = normalized.length > MAX_LOGGED_IDENTITY_CHARS;
-    return {
-        display: normalized.slice(0, MAX_LOGGED_IDENTITY_CHARS),
-        hash: createHash("sha256").update(normalized).digest("hex"),
-        status: truncated ? "truncated" : "complete",
-    };
-}
-/**
- * Classify a turn's prompt-cache shortfall. Returns `null` when no comparable
- * miss exists: no prior readable usage record or a healthy read. The caller
- * separately records why an unreadable/first zero sample could not be
- * classified, so those evidence gaps remain visible in the session summary.
- *
- * Rules, in the order they are applied, each stated with the evidence behind
- * it (2026-08-21 investigation, 63 sessions, 2,288 turns, 94.2% overall hit
- * rate):
- *
- *   1. `prefix-broke` — a `cache_prefix_break` CHANGE fired in this session
- *      since the previous usage record. This is a DIRECT local observation of
- *      the cause, so it outranks the timing heuristics below. The window was
- *      byte-stable throughout, so this verdict should stay rare; if it does
- *      not, #1016's fix has regressed.
- *   2. `model-provider-changed` — both adjacent identities were present and
- *      differ. This is direct local evidence and outranks timing heuristics.
- *   3. Evidence gates — heuristic causes require a REQUEST-time observation,
- *      a complete bounded sequence hash, adjacent full provider/model identity,
- *      and well-formed provider usage numbers.
- *      Missing evidence returns a specific unknown reason; it never lets a
- *      plausible but unproved heuristic win.
- *   4. `ttl-expired` — a REQUEST-time gap reached the configured provider TTL.
- *      A message-end fallback includes generation time and cannot prove this.
- *   5. `partial-eviction` — a non-zero but low read, where fresh input runs far
- *      above the new content we measured. pi-lens injections average ~2.7
- *      estimated tokens per turn, so they cannot account for a multiple of the
- *      measured content; the remainder is provider-side eviction, typically at
- *      a context limit. Suppressed when the character accumulators were capped,
- *      because a floor estimate would manufacture the verdict.
- *   6. `unknown` — a real shortfall with no local explanation. A bounded
- *      `cacheMissUnknownReason` names the missing or conflicting evidence;
- *      provider behavior is never guessed.
- */
-function classifyCacheMiss(args) {
-    const cacheRead = readableTokenCount(args.cacheRead);
-    if (cacheRead === undefined)
-        return null;
-    // No baseline turn in this session: a first turn reads zero because nothing
-    // is cached yet, not because anything went wrong.
-    const priorCacheRead = readableTokenCount(args.priorCacheRead);
-    if (priorCacheRead === undefined) {
-        return null;
-    }
-    let kind;
-    if (cacheRead === 0) {
-        kind = "zero-read";
-    }
-    else if (priorCacheRead > 0 &&
-        cacheRead < priorCacheRead * LOW_CACHE_READ_RATIO) {
-        kind = "low-read";
-    }
-    else {
-        return null;
-    }
-    // A missing stable session id makes every cross-event baseline ambiguous. Do
-    // not let the shared fallback bucket turn another session's evidence into a
-    // cause claim (#1996 concurrent primary/secondary invariant).
-    if (!args.stableSessionIdentity) {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: "request-correlation-unavailable",
-        };
-    }
-    if (args.prefixBroke)
-        return { kind, cause: "prefix-broke" };
-    if (args.modelProviderChanged) {
-        return { kind, cause: "model-provider-changed" };
-    }
-    // TTL and transcript-growth claims need the request-side context observation.
-    // A response-time fallback includes generation and can only explain why the
-    // verdict stayed unknown; it cannot establish cache expiry.
-    if (args.gapBasis !== "request-time") {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: "request-correlation-unavailable",
-        };
-    }
-    // TTL and eviction are heuristics. They are only evidence-backed if the
-    // bounded request observation is complete and both adjacent usage records
-    // identify the same provider/model. Otherwise an unseen prompt or identity
-    // change remains a competing explanation, so classification must fail closed.
-    if (args.requestEvidenceIncomplete) {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: "request-evidence-incomplete",
-        };
-    }
-    if (args.sequenceHashTruncated) {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: "sequence-hash-truncated",
-        };
-    }
-    if (args.providerUsageMalformed) {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: "malformed-provider-usage",
-        };
-    }
-    if (!args.modelProviderEvidenceAvailable) {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: "model-provider-unavailable",
-        };
-    }
-    if (args.interTurnGapMs !== null && args.interTurnGapMs >= args.ttlMs) {
-        return { kind, cause: "ttl-expired" };
-    }
-    if (kind === "low-read" &&
-        !args.attributionCharsCapped &&
-        args.input >
-            Math.max(PARTIAL_EVICTION_MIN_FRESH_TOKENS, args.estimatedNewTokens * PARTIAL_EVICTION_INPUT_FACTOR)) {
-        return { kind, cause: "partial-eviction" };
-    }
-    if (args.prefixStable) {
-        return {
-            kind,
-            cause: "unknown",
-            unknownReason: kind === "zero-read"
-                ? "provider-reported-zero-stable-prefix"
-                : "provider-reported-low-read-stable-prefix",
-        };
-    }
-    return {
-        kind,
-        cause: "unknown",
-        unknownReason: "no-local-explanation",
-    };
-}
-function recordCacheUsageSummary(args) {
-    const summary = args.state.summary;
-    summary.usageRecords += 1;
-    let unknownReasonRecorded = false;
-    if (args.unknownReason === "malformed-provider-usage") {
-        summary.unknownEvidenceReasons["malformed-provider-usage"] += 1;
-        unknownReasonRecorded = true;
-    }
-    if (args.cacheRead === undefined) {
-        if (!unknownReasonRecorded) {
-            summary.unknownEvidenceReasons["cache-read-unavailable"] += 1;
-        }
-        return;
-    }
-    if (args.priorCacheRead === undefined) {
-        if (args.cacheRead === 0) {
-            summary.missObservations += 1;
-            if (!unknownReasonRecorded) {
-                summary.unknownEvidenceReasons["no-prior-sample"] += 1;
-            }
-        }
-        else {
-            summary.cacheHits += 1;
-        }
-        return;
-    }
-    if (!args.verdict) {
-        summary.cacheHits += 1;
-        return;
-    }
-    summary.missObservations += 1;
-    summary.cacheMissCauses[args.verdict.cause] += 1;
-    if (args.verdict.cause === "unknown" &&
-        args.unknownReason &&
-        !unknownReasonRecorded) {
-        summary.unknownEvidenceReasons[args.unknownReason] += 1;
-    }
-}
-function boundedHashScalar(value, state) {
-    if (value === null)
-        return "null";
-    if (value === undefined)
-        return "undefined";
-    if (typeof value === "string") {
-        const truncated = value.length > MAX_HASHED_CONTENT_CHARS;
-        if (truncated)
-            state.contentTruncated = true;
-        const suffix = truncated ? ":content-truncated" : "";
-        return `string:${value.length}:${value.slice(0, MAX_HASHED_CONTENT_CHARS)}${suffix}`;
-    }
-    if (typeof value === "number" ||
-        typeof value === "boolean" ||
-        typeof value === "bigint") {
-        return `${typeof value}:${String(value)}`;
-    }
-    return typeof value;
-}
-function boundedHashArray(value, depth, seen, state) {
-    if (value.length > 12)
-        state.contentTruncated = true;
-    const items = value
-        .slice(0, 12)
-        .map((item) => boundedHashValue(item, depth + 1, seen, state));
-    return `array:${value.length}:[${items.join(",")}]`;
-}
-function boundedHashObject(value, depth, seen, state) {
-    const keys = Object.keys(value).sort((left, right) => compareOrdinal(left, right));
-    if (keys.length > 24)
-        state.contentTruncated = true;
-    const fields = keys
-        .slice(0, 24)
-        .map((key) => `${key}:${boundedHashValue(value[key], depth + 1, seen, state)}`);
-    return `object:${keys.length}:{${fields.join(",")}}`;
-}
-function boundedHashValue(value, depth = 0, seen = new Set(), state) {
-    const hashState = state ?? { contentTruncated: false, incomplete: false };
-    if (value === null || typeof value !== "object") {
-        return boundedHashScalar(value, hashState);
-    }
-    if (depth >= 3) {
-        hashState.contentTruncated = true;
-        return "[depth-limit]";
-    }
-    if (seen.has(value)) {
-        hashState.incomplete = true;
-        return "[cycle]";
-    }
-    seen.add(value);
-    try {
-        if (Array.isArray(value)) {
-            return boundedHashArray(value, depth, seen, hashState);
-        }
-        return boundedHashObject(value, depth, seen, hashState);
-    }
-    catch {
-        hashState.incomplete = true;
-        return "[unreadable]";
-    }
-    finally {
-        seen.delete(value);
-    }
-}
-function hashMessageSequence(messages) {
-    const hash = createHash("sha256");
-    const state = {
-        contentTruncated: false,
-        incomplete: false,
-    };
-    let messageCount;
-    try {
-        messageCount = messages.length;
-        if (!Number.isSafeInteger(messageCount) || messageCount < 0) {
-            throw new Error("invalid message count");
-        }
-    }
-    catch {
-        state.incomplete = true;
-        hash.update("message-count:[unreadable];");
-        return {
-            hash: hash.digest("hex"),
-            truncated: false,
-            contentTruncated: false,
-            incomplete: true,
-        };
-    }
-    hash.update(`message-count:${messageCount};`);
-    const sampled = Math.min(messageCount, MAX_HASHED_MESSAGES);
-    for (let i = 0; i < sampled; i++) {
-        const seen = new Set();
-        let role;
-        let content;
-        try {
-            const message = messages[i];
-            role = message?.role;
-            content = message?.content;
-        }
-        catch {
-            state.incomplete = true;
-            role = "[unreadable]";
-            content = "[unreadable]";
-        }
-        hash.update(`${i}|role:${boundedHashValue(role, 0, seen, state)}|content:${boundedHashValue(content, 0, seen, state)};`);
-    }
-    const truncated = messageCount > sampled;
-    if (truncated)
-        hash.update(`truncated-after:${sampled};`);
-    return {
-        hash: hash.digest("hex"),
-        truncated,
-        contentTruncated: state.contentTruncated,
-        incomplete: state.incomplete,
-    };
-}
-/** Bounds for reporting an injected payload's size in the record. */
-const INJECTED_LIMITS = {
-    maxChars: MAX_INJECTED_CHARS,
-    maxBytes: MAX_INJECTED_BYTES,
-};
-/** Bounds for the transcript-growth accumulator that feeds the verdict. */
-const TRANSCRIPT_LIMITS = {
-    maxChars: MAX_ATTRIBUTION_CHARS,
-    maxBytes: MAX_ATTRIBUTION_BYTES,
-};
-function messageTextSize(content, limits) {
-    if (typeof content === "string") {
-        const chars = Math.min(content.length, limits.maxChars);
-        return {
-            chars,
-            bytes: Math.min(Buffer.byteLength(content.slice(0, chars), "utf8"), limits.maxBytes),
-        };
-    }
-    const bounded = boundedHashValue(content, 0, new Set(), {
-        contentTruncated: false,
-        incomplete: false,
-    });
-    return {
-        chars: Math.min(bounded.length, limits.maxChars),
-        bytes: Math.min(Buffer.byteLength(bounded.slice(0, limits.maxChars), "utf8"), limits.maxBytes),
-    };
-}
-function measureMessages(messages, limits) {
-    let chars = 0;
-    let bytes = 0;
-    const measuredCount = Math.min(messages.length, MAX_REPORTED_MESSAGES);
-    for (let i = 0; i < measuredCount; i++) {
-        const message = messages[i];
-        const size = messageTextSize(message?.content, limits);
-        chars = Math.min(limits.maxChars, chars + size.chars);
-        bytes = Math.min(limits.maxBytes, bytes + size.bytes);
-        if (chars === limits.maxChars || bytes === limits.maxBytes)
-            break;
-    }
-    return {
-        chars,
-        bytes,
-        capped: chars >= limits.maxChars || bytes >= limits.maxBytes,
-    };
-}
-/** Size of an injected payload, bounded for reporting. */
-function measureInjectedMessages(messages) {
-    return measureMessages(messages, INJECTED_LIMITS);
-}
-/**
- * Size of the transcript's own growth, bounded far higher. A routine tool
- * result is an order of magnitude past the injection cap, and measuring it
- * through that cap made `partial-eviction` near-unreachable (#1071 review
- * round 1, F2).
- */
-function measureTranscriptGrowth(messages) {
-    return measureMessages(messages, TRANSCRIPT_LIMITS);
-}
-function sessionKey(sessionId) {
-    return sessionId?.trim() ? sessionId.trim() : NO_SESSION_KEY;
-}
-/**
- * A stable host session id is the only safe cross-event identity. When it is
- * absent, keep primary and concurrent-secondary observations in separate
- * buckets and let classification fail closed instead of allowing a secondary
- * to overwrite the primary's baseline.
- */
-function attributionKey(sessionId, sessionRole) {
-    const key = sessionKey(sessionId);
-    if (key !== NO_SESSION_KEY)
-        return key;
-    return `${sessionRole ?? "unknown"}:${key}`;
-}
-/**
- * Log one bounded request-side observation for every `context` call. This is
- * deliberately a separate phase from `cache_prefix_break`: it describes the
- * messages pi-lens saw and returned from its own context handler, not the final
- * provider request after other context handlers and not whether a provider
- * reused or missed a prompt cache. `observationId` is local to this process; the
- * host currently exposes no request id on ContextEvent/MessageEndEvent. A
- * RuntimeCoordinator turn is process-global, so it is explicitly scoped as
- * such and is omitted for concurrent secondary sessions.
- */
-export function observeCacheContext(args) {
-    try {
-        const existingMessages = args.existingMessages ?? [];
-        const resultMessages = args.resultMessages ?? existingMessages;
-        // Drop empty slices before deriving anything: a source that contributed
-        // no message is not a contributor, and listing it would overstate the
-        // payload's provenance.
-        const injectionSlices = (args.injectionSlices ?? []).filter((slice) => slice.messages.length > 0);
-        const injectedMessages = injectionSlices.flatMap((slice) => slice.messages);
-        const placement = args.placement ?? "none";
-        const beforeSequence = hashMessageSequence(existingMessages);
-        const afterSequence = hashMessageSequence(resultMessages);
-        const sizes = measureInjectedMessages(injectedMessages);
-        // Per-source split of a mixed payload (#1071 AC 2). The old record named
-        // which sources fired but attributed cost by CALL, so a turn carrying
-        // turn-findings plus an agent nudge was one undivided number.
-        const sourceBreakdown = injectionSlices.map((slice) => {
-            const sliceSize = measureInjectedMessages(slice.messages);
-            return {
-                source: slice.source,
-                messageCount: Math.min(slice.messages.length, MAX_REPORTED_MESSAGES),
-                chars: sliceSize.chars,
-                bytes: sliceSize.bytes,
-                estimatedTokens: estimateTokens(sliceSize.chars),
-                countsCapped: sliceSize.capped,
-            };
-        });
-        const state = attributionFor(attributionKey(args.sessionId, args.sessionRole));
-        for (const slice of injectionSlices) {
-            const key = sourceKey(slice.source);
-            const bytes = measureInjectedMessages(slice.messages).bytes;
-            state.injectedBytes[key] += bytes;
-            for (const message of slice.messages) {
-                for (const identity of findingIdentities(typeof message.content === "string" ? message.content : "")) {
-                    if (state.findingIdentities.has(identity)) {
-                        state.injectedFindingsRepeated += 1;
-                        continue;
-                    }
-                    if (state.findingIdentities.size >= MAX_FINDING_IDENTITIES) {
-                        if (!state.findingIdentityCapRecorded) {
-                            state.findingIdentityCapRecorded = true;
-                            logLatency({
-                                type: "phase",
-                                filePath: "<pi-lens>",
-                                phase: "injected_finding_identity_cap",
-                                durationMs: 0,
-                                metadata: { cap: MAX_FINDING_IDENTITIES },
-                            });
-                        }
-                        continue;
-                    }
-                    state.findingIdentities.add(identity);
-                }
-            }
-        }
-        const messageCountCapped = existingMessages.length > MAX_REPORTED_MESSAGES ||
-            resultMessages.length > MAX_REPORTED_MESSAGES;
-        recordContextAttribution(attributionKey(args.sessionId, args.sessionRole), existingMessages, sizes);
-        state.sequenceHashTruncatedSinceLastUsage ||=
-            beforeSequence.truncated ||
-                beforeSequence.contentTruncated ||
-                afterSequence.truncated ||
-                afterSequence.contentTruncated;
-        state.requestEvidenceIncompleteSinceLastUsage ||=
-            beforeSequence.incomplete || afterSequence.incomplete;
-        logLatency({
-            type: "phase",
-            filePath: "<pi-lens>",
-            phase: "cache_context",
-            durationMs: 0,
-            metadata: {
-                version: 1,
-                observedStage: "pi-lens-context-handler",
-                observationId: `ctx-${(++contextObservationCounter).toString(36)}`,
-                sessionId: sessionKey(args.sessionId),
-                sessionRole: args.sessionRole,
-                ...(args.sessionRole === "concurrent-secondary"
-                    ? { turnScope: SECONDARY_TURN_SCOPE }
-                    : { turnIndex: args.turnIndex, turnScope: PROCESS_TURN_SCOPE }),
-                injectionEnabled: args.injectionEnabled,
-                injectionSources: sourceBreakdown.map((entry) => entry.source),
-                injectionSourceBreakdown: sourceBreakdown,
-                injectionOccurred: injectedMessages.length > 0,
-                injectedMessageCount: Math.min(injectedMessages.length, MAX_REPORTED_MESSAGES),
-                injectedMessageCountCapped: injectedMessages.length > MAX_REPORTED_MESSAGES,
-                injectedChars: sizes.chars,
-                injectedBytes: sizes.bytes,
-                injectedBytesBySource: Object.fromEntries(sourceBreakdown.map((entry) => [
-                    sourceKey(entry.source),
-                    entry.bytes,
-                ])),
-                injectedEstimatedTokens: estimateTokens(sizes.chars),
-                injectedTokenBasis: "chars-per-token-4-estimate-not-provider-measured",
-                injectedCountsCapped: sizes.capped,
-                existingMessageCount: Math.min(existingMessages.length, MAX_REPORTED_MESSAGES),
-                resultMessageCount: Math.min(resultMessages.length, MAX_REPORTED_MESSAGES),
-                messageCountCapped,
-                placement,
-                beforeSequenceHash: beforeSequence.hash,
-                afterSequenceHash: afterSequence.hash,
-                sequenceHashTruncated: beforeSequence.truncated ||
-                    afterSequence.truncated ||
-                    beforeSequence.contentTruncated ||
-                    afterSequence.contentTruncated,
-                sequenceMessageCountTruncated: beforeSequence.truncated || afterSequence.truncated,
-                sequenceContentHashTruncated: beforeSequence.contentTruncated || afterSequence.contentTruncated,
-                sequenceHashIncomplete: beforeSequence.incomplete || afterSequence.incomplete,
-            },
-        });
-    }
-    catch (err) {
-        // The context event still occurred at request time. Preserve that clock but
-        // fail the evidence closed so a later usage row cannot infer TTL/eviction
-        // from an observation whose hashing/measurement never completed.
-        try {
-            const state = attributionFor(attributionKey(args.sessionId, args.sessionRole));
-            state.lastContextAtMs = Date.now();
-            state.requestEvidenceIncompleteSinceLastUsage = true;
-        }
-        catch {
-            // The outer contract is no-throw; there is no safe attribution key left.
-        }
-        args.dbg?.(`cache-context: failed to log context observation: ${err}`);
-    }
-}
-/**
- * Part 1 — log one `cache_usage` record for an assistant `message_end` that
- * carries a `usage`. Provider/model come straight off the message itself
- * (`AssistantMessage.provider` / `.model` in pi-ai) — no dependency on the
- * runtime telemetry identity. Skips silently (no record) when the message is
- * not an assistant message or has no usage; never logs a zeros-only record for
- * a message that simply lacks usage.
- */
-export function logCacheUsage(message, dbg, context) {
-    try {
-        if (!message || typeof message !== "object")
-            return;
-        const msg = message;
-        // Only assistant messages carry LLM usage; tool-result / user messages
-        // (and unknown custom AgentMessage variants) are skipped.
-        if (msg.role !== "assistant")
-            return;
-        const usage = msg.usage;
-        if (!usage || typeof usage !== "object")
-            return;
-        // SAFETY: `usage` is verified non-null object above; widening to an
-        // indexable record is only so the numeric-evidence readers can probe
-        // each field defensively. Every read re-validates its own type.
-        const usageRecord = usage;
-        // #1071: attribute the shortfall in-process, from state this module already
-        // keeps, so nothing re-reads latency.log.
-        const stateKey = attributionKey(context?.sessionId, context?.sessionRole);
-        const state = attributionFor(stateKey);
-        const nowMs = Date.now();
-        const gap = resolveInterTurnGap(state, nowMs);
-        const interTurnGapMs = gap.gapMs;
-        const ttlMs = getProviderCacheTtlMs();
-        const inputEvidence = readNumericEvidence(usageRecord, "input");
-        const outputEvidence = readNumericEvidence(usageRecord, "output");
-        const cacheReadEvidence = readNumericEvidence(usageRecord, "cacheRead");
-        const cacheWriteEvidence = readNumericEvidence(usageRecord, "cacheWrite");
-        let costEvidence;
-        try {
-            const cost = usageRecord.cost;
-            costEvidence =
-                cost === undefined
-                    ? { status: "absent" }
-                    : cost && typeof cost === "object"
-                        ? readNumericEvidence(cost, "total")
-                        : { status: "malformed" };
-        }
-        catch {
-            costEvidence = { status: "malformed" };
-        }
-        const usageEvidenceFields = [
-            ["input", inputEvidence],
-            ["output", outputEvidence],
-            ["cacheRead", cacheReadEvidence],
-            ["cacheWrite", cacheWriteEvidence],
-            ["cost.total", costEvidence],
-        ];
-        const malformedProviderUsageFields = usageEvidenceFields
-            .filter(([, evidence]) => evidence.status === "malformed")
-            .map(([field]) => field);
-        const providerUsageMalformed = malformedProviderUsageFields.length > 0;
-        const cacheRead = cacheReadEvidence.value;
-        const priorCacheReadValue = readableTokenCount(state.lastCacheRead);
-        // SAFETY: `msg` passed the AssistantMessageLike role check above; the
-        // record view backs only the guarded identity-field probes below, each
-        // re-validated by `identityEvidence` at its own boundary.
-        const messageRecord = msg;
-        const safeMessageField = (key) => {
-            try {
-                return identityEvidence(messageRecord[key]);
-            }
-            catch {
-                return { status: "unavailable" };
-            }
-        };
-        const currentProvider = safeMessageField("provider");
-        const responseModel = safeMessageField("responseModel");
-        const currentModel = responseModel.hash !== undefined
-            ? responseModel
-            : safeMessageField("model");
-        const modelProviderEvidenceAvailable = Boolean(currentProvider.hash &&
-            currentModel.hash &&
-            state.lastProviderIdentityHash &&
-            state.lastModelIdentityHash);
-        const modelProviderChanged = Boolean(modelProviderEvidenceAvailable &&
-            (currentProvider.hash !== state.lastProviderIdentityHash ||
-                currentModel.hash !== state.lastModelIdentityHash));
-        const estimatedNewTokens = estimateTokens(state.injectedCharsSinceLastUsage +
-            state.newTranscriptCharsSinceLastUsage);
-        const verdict = classifyCacheMiss({
-            cacheRead,
-            // Missing/malformed input cannot prove fresh-input pressure. The explicit
-            // malformed gate below also keeps every heuristic fail-closed.
-            input: inputEvidence.value ?? 0,
-            priorCacheRead: priorCacheReadValue,
-            interTurnGapMs,
-            gapBasis: gap.basis,
-            ttlMs,
-            prefixBroke: state.prefixBrokeSinceLastUsage,
-            prefixStable: state.prefixStableSinceLastUsage,
-            sequenceHashTruncated: state.sequenceHashTruncatedSinceLastUsage,
-            requestEvidenceIncomplete: state.requestEvidenceIncompleteSinceLastUsage,
-            stableSessionIdentity: sessionKey(context?.sessionId) !== NO_SESSION_KEY,
-            modelProviderChanged,
-            modelProviderEvidenceAvailable,
-            providerUsageMalformed,
-            estimatedNewTokens,
-            attributionCharsCapped: state.attributionCharsCapped,
-        });
-        const unknownReason = providerUsageMalformed
-            ? "malformed-provider-usage"
-            : cacheRead === undefined
-                ? "cache-read-unavailable"
-                : priorCacheReadValue === undefined && cacheRead === 0
-                    ? "no-prior-sample"
-                    : verdict?.unknownReason;
-        recordCacheUsageSummary({
-            state,
-            cacheRead,
-            priorCacheRead: priorCacheReadValue,
-            verdict,
-            unknownReason,
-        });
-        const priorCacheRead = priorCacheReadValue ?? null;
-        const injectedCharsSinceLastTurn = state.injectedCharsSinceLastUsage;
-        const newTranscriptCharsSinceLastTurn = state.newTranscriptCharsSinceLastUsage;
-        const attributionCharsCapped = state.attributionCharsCapped;
-        const injectedBytes = {
-            ...state.injectedBytes,
-            turnEndAdvisory: state.turnEndAdvisoryBytes + state.injectedBytes.turnEndAdvisory,
-        };
-        const injectedFindingsRepeated = state.injectedFindingsRepeated;
-        const toolResultBytes = state.toolResultBytes;
-        const toolResultsTruncated = state.toolResultsTruncated;
-        // This record is the turn boundary: reset the per-turn accumulators and
-        // re-arm the prefix-break flag so the next verdict describes the NEXT gap.
-        state.lastUsageAtMs = nowMs;
-        // A record whose `cacheRead` is absent or non-finite carries no baseline.
-        // Clear the stored one rather than keeping it: the next turn's gap would be
-        // one turn long while its baseline was two turns old, and a verdict built
-        // on that mismatch is worse than no verdict (#1071 review round 1, F4).
-        state.lastCacheRead = cacheRead;
-        state.lastProviderIdentityHash = currentProvider.hash;
-        state.lastModelIdentityHash = currentModel.hash;
-        state.prefixBrokeSinceLastUsage = false;
-        state.prefixStableSinceLastUsage = false;
-        state.sequenceHashTruncatedSinceLastUsage = false;
-        state.requestEvidenceIncompleteSinceLastUsage = false;
-        state.injectedCharsSinceLastUsage = 0;
-        state.newTranscriptCharsSinceLastUsage = 0;
-        state.attributionCharsCapped = false;
-        state.injectedBytes = emptyInjectedBytes();
-        state.injectedFindingsRepeated = 0;
-        state.turnEndAdvisoryBytes = 0;
-        state.toolResultBytes = 0;
-        state.toolResultsTruncated = 0;
-        // Clear the request stamp too: the next turn measures from ITS request, and
-        // a turn whose `context` call pi-lens never saw must fall back rather than
-        // reuse this one.
-        state.lastContextAtMs = undefined;
-        logLatency({
-            type: "phase",
-            filePath: "<pi-lens>",
-            phase: "cache_usage",
-            durationMs: 0,
-            metadata: {
-                provider: currentProvider.display,
-                model: currentModel.display,
-                providerIdentityHash: currentProvider.hash,
-                modelIdentityHash: currentModel.hash,
-                providerIdentityStatus: currentProvider.status,
-                modelIdentityStatus: currentModel.status,
-                providerIdentityTruncated: currentProvider.status === "truncated",
-                modelIdentityTruncated: currentModel.status === "truncated",
-                cacheRead: cacheReadEvidence.value ?? null,
-                cacheWrite: cacheWriteEvidence.value ?? null,
-                input: inputEvidence.value ?? null,
-                output: outputEvidence.value ?? null,
-                // `Usage.cost` is a breakdown object; the total is the headline number.
-                cost: costEvidence.value ?? null,
-                providerUsageMalformedFields: malformedProviderUsageFields,
-                // Idle milliseconds before this turn's request, or null for the first
-                // record in the session. This is the field the 2026-08-21
-                // investigation had to reconstruct by hand-joining timestamps.
-                interTurnGapMs,
-                gapBasis: gap.basis,
-                // null means "nothing to explain", not "cause unknown"; `unknown` is
-                // the explicit no-local-explanation verdict.
-                cacheMissCause: verdict?.cause ?? null,
-                cacheMissKind: verdict?.kind ?? null,
-                cacheMissUnknownReason: unknownReason ?? null,
-                modelProviderChanged,
-                cacheTtlThresholdMs: ttlMs,
-                priorCacheRead,
-                injectedCharsSinceLastTurn,
-                newTranscriptCharsSinceLastTurn,
-                attributionCharsCapped,
-                injectedBytes,
-                injectedFindingsRepeated,
-                // Delivered tool-result aggregation over the turn (#2800 item 7):
-                // the sum of each delivered footer's bytes= figure and the count
-                // of results whose bound fired.
-                toolResultBytes,
-                toolResultsTruncated,
-                ...(context
-                    ? {
-                        // MessageEndEvent has no request/context id in the host API. These
-                        // fields permit session correlation without inventing an exact
-                        // provider-request linkage. The process-global turn is omitted
-                        // for a concurrent secondary session.
-                        sessionId: sessionKey(context.sessionId),
-                        ...(context.sessionRole === "concurrent-secondary"
-                            ? { turnScope: SECONDARY_TURN_SCOPE }
-                            : {
-                                turnIndex: context.turnIndex,
-                                turnScope: PROCESS_TURN_SCOPE,
-                            }),
-                        contextCorrelation: context.sessionId
-                            ? "session-only-no-request-id"
-                            : "no-stable-session-id",
-                    }
-                    : {}),
-            },
-        });
-    }
-    catch (err) {
-        dbg?.(`cache-usage: failed to log message_end usage: ${err}`);
-    }
-}
-/**
- * Content hash of the first transcript message. Stable for identical content
- * (role + content are serialized in a fixed order), so a changing hash means
- * `messages[0]` actually changed byte-for-byte.
- */
-function hashFirstMessage(first) {
-    // Keep the original full local hash semantics for the existing
-    // `cache_prefix_break` signal. The new `cache_context` hashes above are the
-    // bounded hashes; this signal must not silently miss a suffix-only change.
-    const serialized = JSON.stringify({
-        role: first.role,
-        content: first.content,
-    });
-    return createHash("sha256").update(serialized).digest("hex");
-}
-/**
- * Per-session baseline hash of `messages[0]`, keyed by the STABLE pi session id
- * (`ctx.sessionManager.getSessionId()`). A single module-scoped var was wrong:
- * in ONE process, multiple logical conversations touch this module, and they
- * must not share a baseline —
- *   - new / fork / a concurrent in-process subagent (#473) is a DIFFERENT id →
- *     its OWN independent baseline, never compared against another id's (else a
- *     benign session boundary logs a spurious `cache_prefix_break`, and a
- *     concurrent subagent + its parent stomp each other's baseline, each
- *     emitting a false positive — fatal for a signal whose value is trust);
- *   - an IN-PROCESS reload / resume reuses the SAME id → the baseline is still in
- *     this map, so a genuine break is caught (no blinding reset). NOTE: a full
- *     process restart (quit → `pi --session <id>`) starts a NEW process with an
- *     empty map, so the first post-restart `context` re-anchors a fresh baseline
- *     rather than comparing across the restart. That's acceptable here — this is a
- *     pure observability signal (at worst a missed `cache_prefix_break` log on the
- *     first post-restart turn, never a user-facing action) — unlike genuine
- *     session state (read guard #1041, widget #190) which IS rehydrated from the
- *     sidecar.
- *
- * Bounded as insertion-ordered FIFO (evict oldest-inserted past the cap,
- * refreshed on every WRITE by {@link recordSessionHash}'s delete+set — a
- * plain read never reorders) so a long-lived process cycling through many
- * sessions can't grow this unbounded. Shares {@link MAX_TRACKED_SESSIONS}
- * with `attributionBySession` above.
- */
-const prefixHashBySession = new BoundedFifoMap(MAX_TRACKED_SESSIONS);
-/**
- * Reported key used when no session id is available (undefined/empty). Internal
- * attribution additionally separates primary and concurrent-secondary roles;
- * neither role may overwrite the other's baseline when stable identity is
- * absent.
- */
-const NO_SESSION_KEY = "<no-session>";
-/**
- * Store `hash` for `key`, refreshing its LRU recency (re-insert moves it to the
- * newest position since `Map` preserves insertion order) and evicting the
- * oldest-inserted entries once the cap is exceeded.
- */
-function recordSessionHash(key, hash) {
-    prefixHashBySession.delete(key);
-    prefixHashBySession.set(key, hash);
-}
-/**
- * Part 2 — observe `messages[0]` stability turn-over-turn, keyed by session id.
- * Logs a baseline the first time it sees a non-empty transcript FOR A GIVEN
- * session id, then logs a `cache_prefix_break` whenever that same session's
- * first-message hash changes. Pure observation: it never inspects or mutates
- * anything but its own per-session hash map, and never throws.
- *
- * `sessionId` should be pi's STABLE id (`ctx.sessionManager.getSessionId()`),
- * which uniquely identifies the CURRENTLY-firing session: a concurrent
- * in-process subagent runs its own `AgentSession`/`sessionManager`, so its
- * `context` calls carry a DIFFERENT id than the parent's and get their own
- * baseline. (`runtime.telemetrySessionId` cannot be used here: per the #473
- * guard a concurrent secondary skips `updateRuntimeIdentityFromEvent`, so that
- * process-global singleton stays pinned to the PARENT — it would collapse
- * parent and subagent onto one baseline.) If the host does NOT supply a stable
- * id, primary and secondary roles still use separate fallback buckets, but
- * multiple same-role sessions cannot be correlated. Miss attribution therefore
- * fails closed with `request-correlation-unavailable`.
- *
- * Does nothing on an empty transcript (nothing to anchor a prefix to yet).
- */
-export function observeCachePrefix(messages, turnIndex, sessionId, sessionRole, dbg) {
-    try {
-        if (!messages || messages.length === 0)
-            return "empty";
-        const first = messages[0];
-        if (!first || typeof first !== "object")
-            return "empty";
-        const key = attributionKey(sessionId, sessionRole);
-        const reportedSessionId = sessionKey(sessionId);
-        const currentHash = hashFirstMessage(first);
-        const previousHash = prefixHashBySession.get(key);
-        const state = attributionFor(key);
-        if (previousHash === undefined) {
-            // Baseline: record this session's starting prefix so a later break has a
-            // reference point in the log. `previousHash: null` marks the baseline.
-            recordSessionHash(key, currentHash);
-            state.prefixStableSinceLastUsage = true;
-            logLatency({
-                type: "phase",
-                filePath: "<pi-lens>",
-                phase: "cache_prefix_break",
-                durationMs: 0,
-                metadata: {
-                    ...(sessionRole === "concurrent-secondary"
-                        ? { turnScope: SECONDARY_TURN_SCOPE }
-                        : { turnIndex, turnScope: PROCESS_TURN_SCOPE }),
-                    previousHash: null,
-                    currentHash,
-                    baseline: true,
-                    sessionId: reportedSessionId,
-                    sessionRole,
-                },
-            });
-            return "baseline";
-        }
-        if (currentHash !== previousHash) {
-            // #1071: arm the direct-cause flag the next `cache_usage` verdict reads.
-            // A locally observed prefix change outranks the timing heuristics.
-            state.prefixBrokeSinceLastUsage = true;
-            state.prefixStableSinceLastUsage = false;
-            logLatency({
-                type: "phase",
-                filePath: "<pi-lens>",
-                phase: "cache_prefix_break",
-                durationMs: 0,
-                metadata: {
-                    ...(sessionRole === "concurrent-secondary"
-                        ? { turnScope: SECONDARY_TURN_SCOPE }
-                        : { turnIndex, turnScope: PROCESS_TURN_SCOPE }),
-                    previousHash,
-                    currentHash,
-                    sessionId: reportedSessionId,
-                    sessionRole,
-                },
-            });
-        }
-        if (currentHash === previousHash) {
-            state.prefixStableSinceLastUsage = true;
-        }
-        // Refresh recency (and update the stored hash after a break) so an active
-        // session stays warm in the LRU and isn't evicted while still in use.
-        recordSessionHash(key, currentHash);
-        return currentHash === previousHash ? "unchanged" : "changed";
-    }
-    catch (err) {
-        dbg?.(`cache-prefix: failed to observe messages[0]: ${err}`);
-        return "empty";
-    }
-}
-/**
- * Drop a single session's prefix baseline and attribution state. Called from
- * both role-specific `session_shutdown` paths so an ended conversation is
- * reclaimed promptly rather than only when the LRU evicts it. Idempotent and
- * never throws.
- */
-export function clearCachePrefixSession(sessionId, sessionRole) {
-    const key = attributionKey(sessionId, sessionRole);
-    prefixHashBySession.delete(key);
-    // The miss-attribution state has the same lifetime and the same reason to be
-    // reclaimed at session end; leaving it would let a reused session id inherit
-    // a stale gap baseline and a stale prefix-break flag.
-    attributionBySession.delete(key);
-}
-/** Record the exact quiet-window summary content handed to pi.sendMessage. */
-export function recordTurnEndAdvisoryBytes(sessionId, bytes, sessionRole = "primary") {
-    const state = attributionFor(attributionKey(sessionId, sessionRole));
-    state.turnEndAdvisoryBytes = Math.min(MAX_INJECTED_BYTES, state.turnEndAdvisoryBytes + Math.max(0, bytes));
-}
-/**
- * Fold one delivered tool result into the per-turn aggregation the
- * `cache_usage` row reports as `toolResultBytes` / `toolResultsTruncated`
- * (#2800 item 7). `bytes` is the delivered payload's UTF-8 byte count as
- * stamped in the result's own footer, and `truncated` is the footer's
- * truncated flag. Called from each host adapter's delivery seam; the row is
- * emitted by `logCacheUsage`, which resets both figures at the turn boundary.
- * The byte sum saturates at the attribution bound so the row stays bounded.
- */
-export function recordToolResultDelivery(args) {
-    const state = attributionFor(attributionKey(args.sessionId, args.sessionRole));
-    const bytes = Number.isFinite(args.bytes)
-        ? Math.max(0, Math.floor(args.bytes))
-        : 0;
-    state.toolResultBytes = Math.min(MAX_ATTRIBUTION_BYTES, state.toolResultBytes + bytes);
-    if (args.truncated === true)
-        state.toolResultsTruncated += 1;
-}
-/**
- * Emit one bounded session summary and retire its attribution state. The
- * summary contains fixed-key counters only; it never serializes transcript or
- * context content. A session with no usage records emits nothing.
- */
-export function emitCacheUsageSummaryAtSessionEnd(sessionId, sessionRole) {
-    const key = attributionKey(sessionId, sessionRole);
-    const state = attributionBySession.get(key);
-    if (!state || state.summary.usageRecords === 0)
-        return;
-    logLatency({
-        type: "phase",
-        filePath: "<pi-lens>",
-        phase: "cache_usage_summary",
-        durationMs: 0,
-        metadata: {
-            version: 1,
-            sessionId: sessionKey(sessionId),
-            sessionRole,
-            usageRecords: state.summary.usageRecords,
-            cacheHits: state.summary.cacheHits,
-            missObservations: state.summary.missObservations,
-            cacheMissCauses: { ...state.summary.cacheMissCauses },
-            unknownEvidenceReasons: {
-                ...state.summary.unknownEvidenceReasons,
-            },
-        },
-    });
-    attributionBySession.delete(key);
-}
-/**
- * Clear all per-session prefix hashes and miss-attribution state. For tests and
- * session boundaries. Both maps are cleared together so a test can never start
- * with a half-reset session.
- */
-export function resetCachePrefixObservation() {
-    prefixHashBySession.clear();
-    attributionBySession.clear();
-}
-export function resetCacheFindingIdentitiesSession(sessionId, sessionRole) {
-    const state = attributionBySession.get(attributionKey(sessionId, sessionRole));
-    if (!state)
-        return;
-    state.findingIdentities.clear();
-    state.findingIdentityCapRecorded = false;
-}
-export function _attributionBySessionHasForTests(key) {
-    return attributionBySession.has(key);
-}
-export function _prefixHashBySessionHasForTests(key) {
-    return prefixHashBySession.has(key);
-}
-/** #2442 test-only: drive attributionFor's fetch-or-create touch directly
- *  (every call refreshes LRU recency, matching production — see
- *  attributionFor's own doc comment). */
-export function _touchAttributionForTests(key) {
-    attributionFor(key);
-}
-/** #2442 test-only: drive recordSessionHash's write-refresh directly. */
-export function _recordSessionHashForTests(key, hash) {
-    recordSessionHash(key, hash);
-}
-/**
- * #2442 test-only: the exact `prefixHashBySession.get(key)` that
- * `observeCachePrefix` performs to fetch a session's previous hash. A
- * `.has()` reorders nothing on either bounded class, so only a real `get`
- * can pin FIFO-vs-LRU behavior (#2442 review F4).
- */
-export function _prefixHashBySessionGetForTests(key) {
-    return prefixHashBySession.get(key);
-}