feat(cli): shared hook efficacy ledger + read-after-served adoption KPI

packages/cli/src/repowise/cli/commands/augment_cmd/command.py

@@ -83,6 +83,7 @@
 from .codex import _handle_codex_context_event, _handle_post_edit_use
 from .read_state import _handle_edit_post, _handle_read_post, _record_edit
 from .search import _handle_search_post
+from .served_reads import _handle_mcp_read_post, _log_read_after_served
 from .session_start import _handle_claude_session_start
 
 _EDIT_TOOL_NAMES = {"apply_patch", "Edit", "Write"}
@@ -246,11 +247,18 @@ def _handle_post_tool_use(
             return _handle_post_edit_use(cwd)
         return _handle_edit_post(tool_input, cwd, session_id)
     if tool_name == "Read":
+        # Read-after-served KPI: logged to the ledger, never spoken about.
+        _log_read_after_served(tool_input, tool_output, cwd, session_id)
         return _handle_read_post(tool_input, tool_output, cwd, session_id)
     if tool_name in ("Bash", "PowerShell"):
         # The PowerShell tool (Windows Claude Code) surfaces the same
         # stdout/stderr response shape as Bash — one handler covers both.
         return _handle_bash_post(tool_input, tool_output, cwd)
     if tool_name in ("Grep", "Glob"):
-        return _handle_search_post(tool_name, tool_input, tool_output, cwd)
+        return _handle_search_post(tool_name, tool_input, tool_output, cwd, session_id)
+    if tool_name.startswith("mcp__") and "repowise" in tool_name.lower():
+        # Served-content bookkeeping for the read-after-served KPI. Never
+        # emits — measurement only.
+        _handle_mcp_read_post(tool_output, cwd, session_id)
+        return None
     return None

packages/cli/src/repowise/cli/commands/augment_cmd/decision_inject.py

@@ -547,9 +547,20 @@ def _edit_decision_notice(repo_path: Path, rel: str, session_id: str, state: dic
     "session_id TEXT NOT NULL, decision_id TEXT NOT NULL, "
     "node_id TEXT NOT NULL DEFAULT '', shown_at REAL NOT NULL, "
     "evaluated INTEGER NOT NULL DEFAULT 0, "
+    "surface TEXT NOT NULL DEFAULT '', "
+    "category TEXT NOT NULL DEFAULT '', "
+    "chars INTEGER NOT NULL DEFAULT 0, "
     "PRIMARY KEY (session_id, decision_id))"
 )
 
+#: Mirror of core.sessions.staging.INJECTIONS_LEDGER_COLUMNS — the hook path
+#: must not import repowise.core, so the migration is duplicated verbatim.
+_LEDGER_COLUMNS = (
+    ("surface", "TEXT NOT NULL DEFAULT ''"),
+    ("category", "TEXT NOT NULL DEFAULT ''"),
+    ("chars", "INTEGER NOT NULL DEFAULT 0"),
+)
+
 
 def _open_injections(repo_path: Path) -> sqlite3.Connection | None:
     db_path = repo_path / ".repowise" / "sessions" / "sessions.db"
@@ -559,6 +570,10 @@ def _open_injections(repo_path: Path) -> sqlite3.Connection | None:
         conn.execute("PRAGMA journal_mode=WAL")
         conn.execute("PRAGMA busy_timeout=1000")
         conn.execute(_INJECTIONS_TABLE_SQL)
+        existing = {row[1] for row in conn.execute("PRAGMA table_info(injections)")}
+        for name, decl in _LEDGER_COLUMNS:
+            if name not in existing:
+                conn.execute(f"ALTER TABLE injections ADD COLUMN {name} {decl}")
         return conn
     except (sqlite3.Error, OSError):
         return None
@@ -581,8 +596,9 @@ def _record_injections(
     try:
         now = time.time()
         conn.executemany(
-            "INSERT OR IGNORE INTO injections (session_id, decision_id, node_id, shown_at) "
-            "VALUES (?, ?, ?, ?)",
+            "INSERT OR IGNORE INTO injections "
+            "(session_id, decision_id, node_id, shown_at, surface, category) "
+            "VALUES (?, ?, ?, ?, 'decision', 'session_start')",
             [(session_id, did, node_id, now) for did in decision_ids],
         )
         conn.commit()
@@ -592,6 +608,48 @@ def _record_injections(
         conn.close()
 
 
+def _claim_ledger(
+    repo_path: Path,
+    session_id: str,
+    key: str,
+    *,
+    node_id: str,
+    surface: str,
+    category: str,
+    chars: int,
+) -> tuple[bool, int]:
+    """Atomically claim one non-decision ledger emission.
+
+    Generic twin of :func:`_claim_injection` for the read/search enrichment
+    surfaces: *key* replaces the decision id in the primary key, so INSERT OR
+    IGNORE is the once-per-session-per-key gate. Returns ``(claimed,
+    surface_injection_count)`` where the count covers only rows that actually
+    carried text (``chars > 0``) on *surface* — pure measurement rows must not
+    eat into an injection cap. Fail-closed: any error reports unclaimed.
+    """
+    conn = _open_injections(repo_path)
+    if conn is None:
+        return False, 0
+    try:
+        cur = conn.execute(
+            "INSERT OR IGNORE INTO injections "
+            "(session_id, decision_id, node_id, shown_at, surface, category, chars) "
+            "VALUES (?, ?, ?, ?, ?, ?, ?)",
+            (session_id, key, node_id, time.time(), surface, category, chars),
+        )
+        claimed = cur.rowcount > 0
+        count = conn.execute(
+            "SELECT COUNT(*) FROM injections WHERE session_id = ? AND surface = ? AND chars > 0",
+            (session_id, surface),
+        ).fetchone()[0]
+        conn.commit()
+        return claimed, int(count)
+    except sqlite3.Error:
+        return False, 0
+    finally:
+        conn.close()
+
+
 def _claim_injection(
     repo_path: Path, session_id: str, decision_id: str, node_id: str
 ) -> tuple[bool, int]:
@@ -608,13 +666,17 @@ def _claim_injection(
         return False, 0
     try:
         cur = conn.execute(
-            "INSERT OR IGNORE INTO injections (session_id, decision_id, node_id, shown_at) "
-            "VALUES (?, ?, ?, ?)",
+            "INSERT OR IGNORE INTO injections "
+            "(session_id, decision_id, node_id, shown_at, surface, category) "
+            "VALUES (?, ?, ?, ?, 'decision', 'edit_notice')",
             (session_id, decision_id, node_id, time.time()),
         )
         claimed = cur.rowcount > 0
+        # Surface-scoped: read/search enrichment rows also carry a node_id and
+        # must not eat into the edit-notice cap.
         count = conn.execute(
-            "SELECT COUNT(*) FROM injections WHERE session_id = ? AND node_id != ''",
+            "SELECT COUNT(*) FROM injections WHERE session_id = ? AND node_id != '' "
+            "AND surface IN ('', 'decision')",
             (session_id,),
         ).fetchone()[0]
         conn.commit()

packages/cli/src/repowise/cli/commands/augment_cmd/read_state.py

@@ -50,6 +50,9 @@ def _load_session_state(repo_path: Path, session_id: str) -> dict:
         "stale_notified": [],
         "reread_notified": [],
         "decisions_shown": [],
+        # File ranges served by repowise MCP responses this session, kept for
+        # the read-after-served KPI (see read_enrich; rel -> [[start, end]]).
+        "served": {},
     }
     try:
         state = json.loads(_session_state_path(repo_path).read_text(encoding="utf-8"))

packages/cli/src/repowise/cli/commands/augment_cmd/search.py

@@ -21,6 +21,7 @@ def _handle_search_post(
     tool_input: dict,
     tool_output: object,
     cwd: str,
+    session_id: str = "",
 ) -> str | None:
     """Decide whether to enrich a Grep/Glob result and how."""
     repo_path = _find_repo_root(Path(cwd))
@@ -40,6 +41,7 @@ def _handle_search_post(
     if result_count >= _DIGEST_THRESHOLD:
         digest = _grep_flood_digest(repo_path, output_text)
         if digest:
+            _log_search_firing(repo_path, session_id, "digest", output_text, digest)
             return digest
         # Unparseable output (e.g. Glob path lists): fall through to triage.
 
@@ -69,7 +71,38 @@ def _handle_search_post(
 
     import asyncio
 
-    return asyncio.run(_search_enrich(repo_path, pattern, mode, result_count))
+    enrichment = asyncio.run(_search_enrich(repo_path, pattern, mode, result_count))
+    if enrichment:
+        _log_search_firing(repo_path, session_id, mode, pattern, enrichment)
+    return enrichment
+
+
+def _log_search_firing(
+    repo_path: Path, session_id: str, category: str, keyed_on: str, text: str
+) -> None:
+    """Record one search enrichment in the shared ledger; measurement only.
+
+    All hook surfaces share the sessions.db efficacy ledger so the miner can
+    classify used vs ignored firings in one pass. Keyed on the category plus a
+    content hash — the same rescue repeated in one session logs once. Never
+    changes what the agent sees; any failure is silent.
+    """
+    if not session_id:
+        return
+    import hashlib
+
+    from .decision_inject import _claim_ledger
+
+    digest = hashlib.sha1(keyed_on.encode("utf-8", "replace")).hexdigest()[:12]
+    _claim_ledger(
+        repo_path,
+        session_id,
+        f"search:{category}:{digest}",
+        node_id="",
+        surface="search",
+        category=category,
+        chars=len(text),
+    )
 
 
 def _grep_flood_digest(repo_path: Path, output_text: str) -> str | None: