diff --git a/docs/active-work.md b/docs/active-work.md
index b67334d..a19af89 100644
--- a/docs/active-work.md
+++ b/docs/active-work.md
@@ -524,12 +524,18 @@ has **no branch and no framing yet**.
 ## Test-improvement arc, lane 6 — `scripts/bite` positive control
 
 - Portable branch: `githubsucks/bite-positive-control`, worktree
-  `../pmacs-bite`, based on `main` @ `300cbc4`. First lane of the
-  test-improvement arc scoped by `TEST_IMPROVEMENT.md` §8. Tooling and
+  `../pmacs-bite`, based on `main` @ `300cbc4`. Tooling and
   documentation only — **no product code, no tests changed.**
-- Sequenced **first**, ahead of the arc's own priority list, because
-  every later lane's evidence is bite-shaped. A verifier that cannot
-  fail is not worth more than the claims it certifies.
+- **Self-contained deliberately.** An earlier draft of this entry
+  scoped the lane by citing a `TEST_IMPROVEMENT.md` at the repo root.
+  That file is **untracked and machine-local** — not on `main`, not in
+  this PR's tree — so the citation was a dangling reference for any
+  other machine, which is exactly what this ledger exists to prevent.
+  The lane is described here on its own terms instead. If that audit
+  is to scope later lanes it needs committing first, in its own PR.
+- Sequenced ahead of the rest of the testing work because every later
+  lane's evidence is bite-shaped. A verifier that cannot fail is not
+  worth more than the claims it certifies.
 - **The defect:** the script ran only the swapped tree, so a failing
   swapped run was the only thing it checked. A test that fails
   everywhere — a typo, an unrelated compile break — therefore printed
@@ -556,7 +562,6 @@ has **no branch and no framing yet**.
   `git fetch githubsucks && git worktree add ../pmacs-bite
   -b bite-positive-control githubsucks/bite-positive-control`.
 
-
 ## Parked lane: kill-ring browser + persistence
 
 - Portable branch: `githubsucks/kill-ring-browser`
diff --git a/docs/agent-handoff.md b/docs/agent-handoff.md
index 7c80ee9..5307799 100644
--- a/docs/agent-handoff.md
+++ b/docs/agent-handoff.md
@@ -1597,7 +1597,14 @@ round-trip cannot detect a discriminant shift.
   a commit splits worktree from branch — but do not repeat the
   destroys-your-work claim, which will push the next reader toward
   `git stash` to protect themselves, straight into the repo-global
-  trap above. Corollary for a NEW file: the swap-over-`git show` mode
+  trap above. **The #165 incident itself is now unexplained**, and
+  that is recorded rather than papered over: work really was lost, but
+  not by the mechanism this file blamed. A `SIGKILL` bypasses the trap
+  and would leave the swapped file in place, which is one candidate;
+  so is a stash collision, given the same round. Do not invent a
+  mechanism to close the gap — an unexplained incident is safer than a
+  confident wrong cause, which is what produced this correction.
+  Corollary for a NEW file: the swap-over-`git show` mode
   does not apply at all, so its claims must be bitten by hand-editing.
 - **A CONFLICTING PR silently runs no CI at all.** GitHub builds
   `pull_request` workflow runs against the PR's **merge ref**, which it
diff --git a/scripts/bite b/scripts/bite
index 00b367e..212674f 100755
--- a/scripts/bite
+++ b/scripts/bite
@@ -35,6 +35,16 @@
 # an assertion failure; treat a compile failure as a prompt to narrow
 # the swap to a file that compiles both ways.
 #
+# That classification reads libtest's own `test result:` summary and
+# deliberately does NOT grep for compiler text. A test may PRINT
+# compiler output on failure --- `compile_mode_acceptance` has a
+# fixture that emits `error[E0308]: mismatched types` at column 0, and
+# the compile-mode and editops suites are this script's primary
+# consumers --- so a `^error[E...]` grep would stamp "weaker evidence"
+# on exactly the strong result it was meant to identify. `test result:`
+# is emitted whenever the harness ran, is absent when the swapped file
+# did not build, and stays uncolored when piped.
+#
 # Why this exists: bite-verification is step 4 of the working method,
 # and the obvious shortcut --- git stash --- is a trap here. The stash
 # namespace is REPO-GLOBAL: shared across every worktree and with
@@ -60,6 +70,17 @@ if [ ! -f "$path" ]; then
     exit 2
 fi
 
+# Both cargo runs are CAPTURED rather than streamed, because their
+# output is parsed (passed-counts for the control, `test result:` for
+# the classification). A long acceptance suite therefore prints
+# nothing until it finishes --- that is the trade, not a bug. Do not
+# "fix" it back to streaming without giving the parser another source.
+#
+# Colour is pinned off on both runs: `--color always` in the
+# passed-through args would otherwise wrap the summary lines in escape
+# codes, and every anchored match here would silently stop matching.
+export CARGO_TERM_COLOR=never
+
 # Sum the `N passed` figures across every `test result:` line, so a
 # multi-target invocation is counted correctly rather than only its
 # last binary.
@@ -110,12 +131,23 @@ if [ "$swapped_status" -eq 0 ]; then
     exit 1
 fi
 
-# Distinguish the two ways the swapped run can fail. An assertion
-# failure is the evidence we want; a compile failure only says the old
-# file does not build here.
-if printf '%s\n' "$swapped_out" | grep -q '^error\[E[0-9]*\]\|^error: could not compile'; then
+# Distinguish the ways the swapped run can fail. An assertion failure
+# is the evidence we want; a build failure only says the old file does
+# not compile here, and the tests may never have run at all.
+#
+# Classified from libtest's summary line, NOT from compiler text --- a
+# failing test can print `error[E0308]: ...` at column 0 itself (see
+# the header). `test result: FAILED` means the harness ran and tests
+# failed; no `test result:` line at all means nothing ran.
+if printf '%s\n' "$swapped_out" | grep -q '^test result: FAILED'; then
+    echo "bite: OK (assertion) --- tests fail against $ref:$path (the fix bites)"
+elif printf '%s\n' "$swapped_out" | grep -q '^test result:'; then
+    # Some target ran to completion without a failure, yet cargo still
+    # exited non-zero --- typically a *different* target in the same
+    # invocation failed to build. Neither clean result applies.
+    echo "bite: OK (MIXED) --- some tests ran and none failed, but the run still failed." >&2
+    echo "bite: narrow the invocation to one target before trusting this." >&2
+else
     echo "bite: OK (COMPILE) --- $ref:$path does not build against the current tree." >&2
     echo "bite: weaker evidence than an assertion failure --- the tests may never have run." >&2
-else
-    echo "bite: OK (assertion) --- tests fail against $ref:$path (the fix bites)"
 fi