feat(plugin): the seam that erases the evidence is where the unresolved rules get named (#4216)

Step 5 of milestone 419. The milestone's subject is that a rule read and
ignored is arithmetically identical to a rule read and followed, and the
compaction is where that identity becomes permanent — the turns holding the
evidence are summarised away, and the unjudged thing survives as nothing.

WHY THIS IS ASSEMBLED IN THE HOOK. `rule_usage_events` has no session column;
it is per user over a window. A session-scoped answer therefore cannot be
asked of the server, and has to be built where a session is a thing that
exists. Four ledgers four hooks already write:

  .rules.ids      an arm NAMED the rule
  .opened.ids     the session called get_rule      (#4100)
  .acted.ids      the session called rule_outcome  (new here)
  .checkpoint.ids the rule HELD an act             (#4214)

Every one is an observed tool call. Nothing asks the model what it followed —
milestone 386 ruled that out, because a model asked "did you apply rule 156?"
says yes. Two subtractions: named-minus-opened is the arm talking to nobody,
opened-minus-acted is the milestone's whole subject.

PreCompact stdout is the compaction's custom instructions (#3680), not a
message to the model, so the readout does not say "you slipped" — it says
which ids must be carried through, which is the one thing a summary can do
about an unjudged finding.

SILENT WHEN NOTHING HAPPENED, and the accusations are conditional on having
members. "0 rules unresolved" on every compaction is how a readout teaches
its reader to skip it. Traffic is still reported, because the static
instructions already ask for it in prose; these lines are the measured
version.

scribe_record_outcome.sh is the third ledger's writer, matched on
mcp__.*__rule_outcome and mirroring scribe_record_opened.sh: TMPDIR only,
silent, exit 0 on every path. A PostToolUse hook that spoke would put a line
after every rule_outcome call and give recording an outcome a cost.

Also: check_plugin.py skipped the new hook for want of a smoke event, which
would have left the newest of the three ledgers as the only one the plugin
lane never runs. Added, mirroring its sibling.

tests/test_precompact_hook.py now isolates TMPDIR — the hook reads session
ledgers from there, so without isolation a test would see whatever this real
session had accumulated.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01821k5B3Ysecp9fNYs92Kuy

plugin/hooks/hooks.json

@@ -62,6 +62,15 @@
             "command": "bash \"${CLAUDE_PLUGIN_ROOT}/hooks/scribe_record_opened.sh\""
           }
         ]
+      },
+      {
+        "matcher": "mcp__.*__rule_outcome",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "bash \"${CLAUDE_PLUGIN_ROOT}/hooks/scribe_record_outcome.sh\""
+          }
+        ]
       }
     ],
     "PreCompact": [

plugin/hooks/scribe_defs.sh

@@ -525,6 +525,98 @@ scribe_contract_block() {
   return 0
 }
 
+# ── The session slippage readout (#4216, milestone 419) ───────────────────
+#
+# WHAT IT ANSWERS, AND WHY IT CANNOT BE ASKED OF THE SERVER. Which rules fired
+# this session, which changed an action, which did not. `rule_usage_events`
+# carries no session column — it is per user over a window — so a
+# session-scoped answer has to be assembled where a session is a thing that
+# exists. That is here, from ledgers four hooks already write:
+#
+#   .rules.ids      an arm NAMED the rule           (a teaser was shown)
+#   .opened.ids     the session called get_rule     (#4100, an observed event)
+#   .acted.ids      the session called rule_outcome (#4216)
+#   .checkpoint.ids the rule HELD an act            (#4214, the strongest)
+#
+# EVERY LINE IS AN OBSERVED TOOL CALL. Nothing here asks the model what it
+# followed — milestone 386 ruled that out, and rightly: a model asked "did you
+# apply rule 156?" will say yes. These four files record what HAPPENED.
+#
+# THE SUBTRACTIONS ARE THE POINT. Named-minus-opened is the arm talking to
+# nobody; opened-minus-acted is the milestone's whole subject, a rule read and
+# then indistinguishable from one that worked. Neither is an accusation — a
+# rule may be read and correctly judged not to apply — which is why the lines
+# below ask for the leftovers to be CARRIED, not explained.
+scribe_ledger_ids() {
+  # Live ids from one ledger as space-separated words, for set arithmetic.
+  # Aged through scribe_rules_live so a rule named two hours ago does not read
+  # as something this context still holds; a bare-id ledger (the checkpoint
+  # one) simply has no stamps and survives the ageing unchanged.
+  local f="$1"
+  [ -n "$f" ] && [ -f "$f" ] || return 0
+  scribe_rules_live "$f" | tr ',' ' '
+}
+
+scribe_slippage_lines() {
+  # $1 state dir, $2 sanitised session id.
+  #
+  # SILENT ONLY WHEN NO RULE TOUCHED THE SESSION AT ALL. Traffic that did
+  # happen is always reported, because "which rules governed this work" is
+  # what the static instructions above already ask the summariser to preserve
+  # in prose — these lines are the measured version of that, and they are
+  # three short lines.
+  #
+  # What is conditional is the ACCUSATION. Each subtraction prints only when
+  # it has members, so a session that opened everything it was shown and
+  # resolved everything it opened gets the traffic and no more. "0 rules
+  # unresolved" on every compaction is how a readout teaches its reader to
+  # skip it.
+  local dir="$1" sid="$2" named opened acted held
+  [ -n "$dir" ] && [ -n "$sid" ] || return 0
+  named=$(scribe_ledger_ids "$dir/${sid}.rules.ids")
+  opened=$(scribe_ledger_ids "$dir/${sid}.opened.ids")
+  acted=$(scribe_ledger_ids "$dir/${sid}.acted.ids")
+  held=$(scribe_ledger_ids "$dir/${sid}.checkpoint.ids")
+  [ -n "$named$opened" ] || return 0
+
+  local unread unresolved
+  unread=$(scribe_ids_minus "$named" "$opened")
+  unresolved=$(scribe_ids_minus "$opened" "$acted")
+
+  printf -- '- This session'"'"'s rule traffic, from what actually happened rather than from recollection:\n'
+  [ -n "$opened" ]     && printf -- '  read: %s\n' "$opened"
+  [ -n "$held" ]       && printf -- '  held an act before it ran: %s\n' "$held"
+  [ -n "$unread" ]     && printf -- '  named by an arm and never opened: %s\n' "$unread"
+  if [ -n "$unresolved" ]; then
+    printf -- '  READ WITH NO OUTCOME RECORDED: %s. Carry these over as outstanding. A rule read and left unresolved looks exactly like one that worked, and the summary is where that difference is lost for good — say `rule_outcome(id, "applied")`, or `rule_outcome(id, "departed", why=...)` where you deliberately went another way.\n' "$unresolved"
+  fi
+  return 0
+}
+
+scribe_ids_minus() {
+  # Set difference over two space-separated id lists, order preserved. Written
+  # as one awk pass rather than a nested shell loop because the ledgers can
+  # hold a few dozen ids by the end of a long session and this runs inside the
+  # compaction path, where a slow hook delays the thing it is decorating.
+  local a="$1" b="$2"
+  [ -n "$a" ] || return 0
+  awk -v a="$a" -v b="$b" '
+    BEGIN {
+      n = split(b, drop, " ")
+      for (i = 1; i <= n; i++) if (drop[i] != "") skip[drop[i]] = 1
+      m = split(a, keep, " ")
+      out = ""
+      for (i = 1; i <= m; i++) {
+        id = keep[i]
+        if (id == "" || (id in skip) || (id in done)) continue
+        done[id] = 1
+        out = out (out == "" ? "" : " ") id
+      }
+      if (out != "") print out
+    }
+  ' </dev/null 2>/dev/null
+}
+
 scribe_local_dups() {
   local root="$1" rel="$2" kind name pat hits count label files
   while IFS=$'\t' read -r kind name; do

plugin/hooks/scribe_precompact_preserve.sh

@@ -46,11 +46,22 @@
 # text is the receipt. (Auto-compaction suppresses that notification.)
 set -uo pipefail
 
-# Drain the event so the caller never sees a broken pipe. Nothing in it changes
-# what we emit: the instruction is the same whether the operator typed
-# `/compact` or the session hit its limit, and it composes with any custom
-# instructions they gave, which are merged ahead of ours.
-cat >/dev/null 2>&1 || true
+# shellcheck source=plugin/hooks/scribe_defs.sh
+. "$(dirname "${BASH_SOURCE[0]}")/scribe_defs.sh"
+
+# THE EVENT IS NOW READ, where it used to be drained. The old comment here said
+# nothing in it changes what we emit, and that was true while every line was
+# static. It stopped being true at #4216: the slippage readout below is
+# specific to THIS session, and `session_id` is the key to the ledgers that
+# hold it. The instruction is still the same whether the operator typed
+# `/compact` or the session hit its limit — what varies is the measurement
+# appended to it.
+event=$(cat 2>/dev/null || true)
+session_id=""
+if [ -n "$event" ]; then
+  event_flat=$(printf '%s' "$event" | scribe_json_flat 2>/dev/null || true)
+  session_id=$(scribe_json_pick "$event_flat" '.session_id' 2>/dev/null || true)
+fi
 
 cat <<'EOF'
 Preserve the following literally in the summary — copied through, not
@@ -71,4 +82,29 @@ paraphrased or counted:
 Everything else here can be recovered from the repository or from Scribe. These
 cannot: they are this session's only copy.
 EOF
+
+# ── The slippage readout (#4216, milestone 419) ───────────────────────────
+#
+# WHY IT BELONGS HERE RATHER THAN IN A REPLY. A rule read and left unresolved
+# is invisible by construction: it looks exactly like a rule that worked. The
+# compaction is where that invisibility becomes permanent — the turns holding
+# the evidence are summarised away, and an unjudged thing that survives as
+# nothing is how a decision quietly becomes nobody's.
+#
+# STDOUT HERE IS THE SUMMARISER'S INSTRUCTIONS, not a message to the model
+# (the header records how that was established). So this does not say "you
+# slipped"; it says WHICH ids must be carried through, which is the one thing
+# the summary can do about it. The next turn then reads them in the summary
+# with the work still attached.
+#
+# SILENT WHEN THERE IS NOTHING TO SAY. A session that opened everything it was
+# shown gets no extra lines. "0 rules unresolved" on every compaction is how a
+# readout teaches its reader to skip it.
+if [ -n "$session_id" ]; then
+  safe_sid=$(printf '%s' "$session_id" | tr -c 'A-Za-z0-9._-' '_')
+  slippage=$(scribe_slippage_lines "${TMPDIR:-/tmp}/scribe-priorart" "$safe_sid" 2>/dev/null || true)
+  if [ -n "$slippage" ]; then
+    printf '\n%s\n' "$slippage"
+  fi
+fi
 exit 0

plugin/hooks/scribe_record_outcome.sh

@@ -0,0 +1,57 @@
+#!/usr/bin/env bash
+# Scribe — record that this session said what a rule DID, not merely that it
+# read one (#4216, milestone 419).
+#
+# THE THIRD LEDGER, AND WHY THE TWO THAT EXIST ARE NOT ENOUGH.
+#
+# `.rules.ids` says a rule was NAMED. `.opened.ids` says it was READ (#4100 —
+# a PostToolUse hook watches the `get_rule` call, so it is a recorded event
+# rather than a model's claim about its own context). Neither can say what
+# happened next, and that is the whole of what milestone 419 is about: a rule
+# read and followed and a rule read and forgotten leave identical traces.
+#
+# `rule_outcome` (#4212) is the call that closes that gap, and the server
+# records it. But the server's row carries no session — `rule_usage_events` is
+# per user over a window — so a SESSION-scoped readout cannot be asked of it.
+# The question "which rules changed something in THIS session" has to be
+# answered where a session is a thing that exists, which is here.
+#
+# SAME EVIDENCE CLASS AS `.opened.ids`, deliberately. A tool call happened or
+# it did not, and the harness reports it either way; nothing here asks the
+# model whether it followed anything. That is the distinction milestone 386
+# drew when it ruled out self-report, and this stays on the right side of it.
+#
+# WHAT IT CANNOT SAY: that the rule was followed WELL, or that `applied` was
+# honest. It records that an outcome was declared. The value of that is not the
+# claim itself — it is that the absence of one becomes visible, which is the
+# state nothing could previously name.
+#
+# EXIT 0, ALWAYS. This decorates a ledger; a bookkeeping failure must never
+# turn a successful tool call into a hook error.
+set -uo pipefail
+
+# shellcheck source=plugin/hooks/scribe_defs.sh
+. "$(dirname "${BASH_SOURCE[0]}")/scribe_defs.sh"
+
+event=$(cat 2>/dev/null || true)
+[ -n "$event" ] || exit 0
+
+event_flat=$(printf '%s' "$event" | scribe_json_flat)
+session_id=$(scribe_json_pick "$event_flat" '.session_id')
+[ -n "$session_id" ] || exit 0
+
+# The matcher in hooks.json narrows to the rule_outcome tools, but the server
+# segment of an MCP tool name varies with how the plugin was installed, so the
+# id is read from the field rather than from an assumed tool name — the same
+# reasoning scribe_record_opened.sh gives.
+rule_id=$(scribe_json_pick "$event_flat" '.tool_input.rule_id')
+rule_id=$(printf '%s' "$rule_id" | tr -cd '0-9')
+[ -n "$rule_id" ] || exit 0
+
+state_dir="${TMPDIR:-/tmp}/scribe-priorart"
+mkdir -p "$state_dir" 2>/dev/null || true
+safe_sid=$(printf '%s' "$session_id" | tr -c 'A-Za-z0-9._-' '_')
+
+# Stamped and append-only like its two siblings, so one reader ages them all.
+printf '%s\n' "$rule_id" | scribe_rules_append "$state_dir/${safe_sid}.acted.ids"
+exit 0