@yahyahammad/venastine 0.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (143) hide show
  1. package/.env.example +14 -0
  2. package/LICENSE +201 -0
  3. package/NOTICE +14 -0
  4. package/PRIVACY.md +66 -0
  5. package/README.md +878 -0
  6. package/SECURITY.md +88 -0
  7. package/THIRD_PARTY_NOTICES.md +100 -0
  8. package/agents/builtin/compactor.md +75 -0
  9. package/agents/builtin/explore.md +77 -0
  10. package/agents/builtin/grill-me.md +27 -0
  11. package/agents/builtin/initializer.md +94 -0
  12. package/agents/builtin/pipeline-reviewer.md +100 -0
  13. package/agents/builtin/plan.md +88 -0
  14. package/agents/builtin/review.md +74 -0
  15. package/agents/manager.py +195 -0
  16. package/agents/subagent_tool.py +215 -0
  17. package/agents/tui_commands.py +136 -0
  18. package/bin/venastine.mjs +382 -0
  19. package/config.py +1198 -0
  20. package/core/__init__.py +0 -0
  21. package/core/approval.py +115 -0
  22. package/core/client.py +1548 -0
  23. package/core/compaction.py +959 -0
  24. package/core/config_loader.py +1563 -0
  25. package/core/events.py +61 -0
  26. package/core/interaction.py +328 -0
  27. package/core/loop.py +1353 -0
  28. package/core/memory.py +370 -0
  29. package/core/model_windows.py +175 -0
  30. package/core/pipeline_models.py +176 -0
  31. package/core/reasoning/__init__.py +0 -0
  32. package/core/reasoning/authorization.py +130 -0
  33. package/core/reasoning/base.py +219 -0
  34. package/core/reasoning/confidence_scoring.py +468 -0
  35. package/core/reasoning/events.py +149 -0
  36. package/core/reasoning/json_retry.py +219 -0
  37. package/core/reasoning/orchestrator.py +1382 -0
  38. package/core/reasoning/output_writer.py +286 -0
  39. package/core/reasoning/payload_validation.py +398 -0
  40. package/core/reasoning/pipeline_storage.py +361 -0
  41. package/core/reasoning/review.py +615 -0
  42. package/core/reasoning/scholar.py +434 -0
  43. package/core/reasoning/source_corpus.py +327 -0
  44. package/core/reasoning/source_scoring.py +855 -0
  45. package/core/replay.py +167 -0
  46. package/core/session.py +153 -0
  47. package/core/workspace_trust.py +207 -0
  48. package/credentials.py +180 -0
  49. package/database.py +252 -0
  50. package/env_secrets.py +77 -0
  51. package/json_store.py +123 -0
  52. package/logging_setup.py +185 -0
  53. package/main.py +1959 -0
  54. package/mcp_client/__init__.py +15 -0
  55. package/mcp_client/client.py +499 -0
  56. package/mcp_client/config.py +379 -0
  57. package/mcp_client/registration.py +182 -0
  58. package/memories/manager.py +132 -0
  59. package/memories/tui_commands.py +232 -0
  60. package/package.json +73 -0
  61. package/project_init/config_files.py +278 -0
  62. package/project_init/doc_sets.py +330 -0
  63. package/project_init/generator.py +564 -0
  64. package/project_init/manifest.py +406 -0
  65. package/project_init/tui_commands.py +144 -0
  66. package/prompts/__init__.py +0 -0
  67. package/prompts/annotate.md +14 -0
  68. package/prompts/assumption_audit.md +27 -0
  69. package/prompts/claim_extraction.md +34 -0
  70. package/prompts/completeness.md +18 -0
  71. package/prompts/confidence_tiers.md +23 -0
  72. package/prompts/critic_pass.md +24 -0
  73. package/prompts/final_synthesis.md +13 -0
  74. package/prompts/initial_generation.md +14 -0
  75. package/prompts/preliminary_plan.md +23 -0
  76. package/prompts/revalidate.md +25 -0
  77. package/prompts/revise.md +15 -0
  78. package/prompts/source_grounding.md +66 -0
  79. package/prompts/system_prompts.py +262 -0
  80. package/prompts/universal_system_prompt +7 -0
  81. package/prompts/untrusted_content +1 -0
  82. package/providers.json.example +92 -0
  83. package/pyproject.toml +100 -0
  84. package/requirements.txt +72 -0
  85. package/safety/__init__.py +0 -0
  86. package/safety/policy_enforcement.py +682 -0
  87. package/security/__init__.py +0 -0
  88. package/security/capability.py +177 -0
  89. package/security/permissions.py +190 -0
  90. package/security/posture.py +302 -0
  91. package/security/protected_paths.py +309 -0
  92. package/security/sandbox.py +985 -0
  93. package/skills/builtin/crypto/cryptography-verification.md +66 -0
  94. package/skills/builtin/math/formal-specification.md +67 -0
  95. package/skills/builtin/math/numerical-methods.md +66 -0
  96. package/skills/builtin/math/proof-writing.md +61 -0
  97. package/skills/builtin/research/experiment-design.md +62 -0
  98. package/skills/builtin/research/literature-review.md +65 -0
  99. package/skills/builtin/research/reproducibility.md +55 -0
  100. package/skills/builtin/research/source-evaluation.md +70 -0
  101. package/skills/builtin/research/statistical-inference.md +63 -0
  102. package/skills/builtin/research/technical-writing.md +59 -0
  103. package/skills/builtin/security/cybersecurity-research.md +63 -0
  104. package/skills/builtin/software/code-review.md +61 -0
  105. package/skills/builtin/software/debugging.md +61 -0
  106. package/skills/builtin/software/test-design.md +67 -0
  107. package/skills/manager.py +138 -0
  108. package/skills/tui_commands.py +111 -0
  109. package/storage.py +1005 -0
  110. package/tools/base.py +305 -0
  111. package/tools/builtin/__init__.py +0 -0
  112. package/tools/builtin/_math_common.py +377 -0
  113. package/tools/builtin/_net_common.py +61 -0
  114. package/tools/builtin/arxiv.py +223 -0
  115. package/tools/builtin/ask_user.py +213 -0
  116. package/tools/builtin/discrete_math.py +134 -0
  117. package/tools/builtin/fetch_url.py +153 -0
  118. package/tools/builtin/file_ops.py +412 -0
  119. package/tools/builtin/geometry.py +148 -0
  120. package/tools/builtin/get_time.py +20 -0
  121. package/tools/builtin/linear_algebra.py +155 -0
  122. package/tools/builtin/load_skill.py +66 -0
  123. package/tools/builtin/logic.py +96 -0
  124. package/tools/builtin/pin.py +193 -0
  125. package/tools/builtin/probability_stats.py +228 -0
  126. package/tools/builtin/project_docs.py +405 -0
  127. package/tools/builtin/remember.py +158 -0
  128. package/tools/builtin/shell.py +461 -0
  129. package/tools/builtin/symbolic_math.py +119 -0
  130. package/tools/builtin/todo.py +147 -0
  131. package/tools/builtin/web_search.py +151 -0
  132. package/tools/context.py +136 -0
  133. package/tools/isolation.py +314 -0
  134. package/tools/registry.py +897 -0
  135. package/tui/app.py +3423 -0
  136. package/tui/app.tcss +458 -0
  137. package/tui/commands.py +61 -0
  138. package/tui/diffs.py +265 -0
  139. package/tui/preferences.py +182 -0
  140. package/tui/ravens.py +146 -0
  141. package/tui/screens.py +943 -0
  142. package/tui/themes.py +366 -0
  143. package/tui/widgets.py +1239 -0
@@ -0,0 +1,361 @@
1
+ """
2
+ core/reasoning/pipeline_storage.py
3
+
4
+ Persistence layer for deep-research pipeline runs (ROADMAP §5). A
5
+ PipelineRunRecord row is created at the start of
6
+ stream_deep_research_pipeline(), checkpointed with a data-only
7
+ update_pipeline_run() call after every pass's trace-log line, and
8
+ flipped to a terminal status (complete / failed) at the end. The
9
+ load_pipeline_run() read API returns the stored state as a plain dict
10
+ for inspection and debugging.
11
+
12
+ Schema deliberately mirrors the codebase convention established by
13
+ storage.py's MessageLog: semantically distinct data lives in separate
14
+ typed columns (claims_json / trace_json / coverage_gaps_json /
15
+ final_report), NOT one catch-all blob. The *_json columns hold
16
+ json.dumps'd strings (same pattern as MessageLog.content); final_report
17
+ is plain text. This keeps a serialization problem in one field from
18
+ corrupting the others -- a persistence layer built for crash-resilience
19
+ shouldn't itself be a single point of failure.
20
+
21
+ Tables created here are registered in SQLModel.metadata and built by
22
+ database.create_db_and_tables(), the same startup hook that creates
23
+ ConversationThread / MessageLog.
24
+ """
25
+
26
+ from __future__ import annotations
27
+
28
+ import json
29
+ import logging
30
+ from datetime import datetime, timezone
31
+ from typing import Optional
32
+ from uuid import UUID, uuid4
33
+
34
+ from sqlmodel import Field, Session, SQLModel
35
+
36
+ from database import engine
37
+ # §27: the thread-kind vocabulary lives beside the table that stores it.
38
+ # This import direction is the only one that exists -- storage.py knows
39
+ # nothing about pipeline runs, and classify_legacy_pass_threads() below is
40
+ # where the two meet.
41
+ from storage import THREAD_KIND_CHAT, THREAD_KIND_RESEARCH_PASS
42
+
43
+ logger = logging.getLogger(__name__)
44
+
45
+
46
+ class PipelineRunRecord(SQLModel, table=True):
47
+ """One row per stream_deep_research_pipeline() invocation. `status` is
48
+ 'running' from creation until the wrapping try/except in the
49
+ orchestrator flips it to 'complete' or 'failed'. The four data
50
+ columns carry the run's structured output (or the partial state
51
+ available at failure time) so an operator can inspect what happened
52
+ without a separate trace log."""
53
+
54
+ id: UUID = Field(default_factory=uuid4, primary_key=True)
55
+ user_query: str = Field(index=True)
56
+ status: str = Field(default="running") # "running" | "complete" | "failed"
57
+ claims_json: str = "[]"
58
+ trace_json: str = "[]"
59
+ coverage_gaps_json: str = "[]"
60
+ # §27 (T2): [{"pass": ..., "thread_id": ...}] for this run's pass
61
+ # threads. Persisted rather than left to the output directory alone
62
+ # (where granted_calls and subagent_reviews live) because those are
63
+ # written once, at the end, by a run that finished -- and a run whose
64
+ # passes are worth chasing through their threads is frequently one
65
+ # that did not. Checkpointed like the other data columns, so an
66
+ # abandoned run keeps the ids it had reached.
67
+ pass_threads_json: str = "[]"
68
+ # E13/#77: [{"candidate", "provider_name", "model", "chars"}] per
69
+ # ensemble-Pass-1 SURVIVOR, so a stored claim's
70
+ # asserted_by_candidates names models this row can still identify.
71
+ # Metadata only: the full texts are per-candidate artifacts and pass
72
+ # threads, and update_pipeline_run strips the in-memory `text` key
73
+ # here rather than duplicating bodies into the database.
74
+ # Checkpointed like the other data columns, so an abandoned run keeps
75
+ # what it had reached.
76
+ candidates_json: str = "[]"
77
+ final_report: str = ""
78
+ started_at: datetime = Field(default_factory=lambda: datetime.now(timezone.utc))
79
+ finished_at: Optional[datetime] = None
80
+
81
+
82
+ # ---------------------------------------------------------------------------
83
+ # ---- Public write API ------------------------------------------------------
84
+ # ---------------------------------------------------------------------------
85
+
86
+ def create_pipeline_run(user_query: str) -> UUID:
87
+ """Inserts a 'running' PipelineRunRecord row and returns its id.
88
+ The orchestrator assigns this id to PipelineRun.run_id so later
89
+ updates can find the row."""
90
+ with Session(engine) as session:
91
+ record = PipelineRunRecord(user_query=user_query, status="running")
92
+ session.add(record)
93
+ session.commit()
94
+ session.refresh(record)
95
+ return record.id
96
+
97
+
98
+ def update_pipeline_run(
99
+ run_id: UUID,
100
+ run,
101
+ status: Optional[str] = None,
102
+ ) -> None:
103
+ """Updates an existing row's data columns (claims / trace /
104
+ coverage_gaps / final_report) from the live PipelineRun, and -- only
105
+ when `status` is given -- flips status and sets finished_at on a
106
+ terminal transition. Raises ValueError if run_id doesn't exist (the
107
+ record was never created or was deleted), which is a real bug, not an
108
+ expected path.
109
+
110
+ `status` is Optional by design: the minimal core only ever passes a
111
+ terminal status, but §5 proper's deferred per-pass checkpointing will
112
+ call this for plain data snapshots WITHOUT a status. Defaulting status
113
+ to a value here would silently reset a 'complete'/'failed' record back
114
+ to that default on a data-only update; Optional=None means "don't touch
115
+ status unless told to," which makes that future extension safe by
116
+ construction.
117
+
118
+ Claims are serialized with vars(c), matching every other
119
+ claim-serialization call site in orchestrator.py (3a/3b/5/6a/6c/final
120
+ synthesis). If a nested-dataclass field is ever added to Claim, ALL of
121
+ those sites migrate to asdict(c) together -- see the note on Claim in
122
+ core/reasoning/base.py.
123
+ """
124
+ with Session(engine) as session:
125
+ record = session.get(PipelineRunRecord, run_id)
126
+ if record is None:
127
+ raise ValueError(f"No pipeline run found with id {run_id}")
128
+ if status is not None:
129
+ record.status = status
130
+ if status in ("complete", "failed"):
131
+ record.finished_at = datetime.now(timezone.utc)
132
+ record.claims_json = json.dumps([vars(c) for c in run.claims])
133
+ record.trace_json = json.dumps(run.trace)
134
+ record.coverage_gaps_json = json.dumps(run.coverage_gaps)
135
+ record.pass_threads_json = json.dumps(run.pass_threads)
136
+ # getattr, not run.candidates: this function is called with test
137
+ # doubles and with runs reconstructed elsewhere, and a checkpoint
138
+ # that raises on a missing optional field would take down the pass
139
+ # that was only trying to record its progress.
140
+ # `text` rides the in-memory entry (it feeds the artifacts and
141
+ # Pass 3b) and is stripped HERE -- the database carries metadata
142
+ # only; see PipelineRunRecord's column comment.
143
+ record.candidates_json = json.dumps([
144
+ {k: v for k, v in entry.items() if k != "text"}
145
+ for entry in getattr(run, "candidates", []) or []
146
+ ])
147
+ record.final_report = run.final_report
148
+ session.add(record)
149
+ session.commit()
150
+
151
+
152
+ # ---------------------------------------------------------------------------
153
+ # ---- Public read API -------------------------------------------------------
154
+ # ---------------------------------------------------------------------------
155
+
156
+ def load_pipeline_run(run_id: UUID) -> dict:
157
+ """Loads a PipelineRunRecord by id and returns its state as a plain
158
+ dict. Raises ValueError if run_id doesn't exist (matching
159
+ update_pipeline_run's convention -- a missing record is a real bug,
160
+ not an expected path).
161
+
162
+ Claims are returned as raw dicts (the shape vars(c) produced at write
163
+ time), not reconstructed Claim dataclass instances. If/when
164
+ programmatic pipeline resume is built, that function reconstructs
165
+ Claim objects itself, using the same field-filtering pattern
166
+ _claim_from_json already established in orchestrator.py.
167
+ """
168
+ with Session(engine) as session:
169
+ record = session.get(PipelineRunRecord, run_id)
170
+ if record is None:
171
+ raise ValueError(f"No pipeline run found with id {run_id}")
172
+ return {
173
+ "id": record.id,
174
+ "user_query": record.user_query,
175
+ "status": record.status,
176
+ "started_at": record.started_at,
177
+ "finished_at": record.finished_at,
178
+ "claims": json.loads(record.claims_json),
179
+ "trace": json.loads(record.trace_json),
180
+ "coverage_gaps": json.loads(record.coverage_gaps_json),
181
+ # §27 AC6. Defaulted rather than read straight off the record:
182
+ # a row written before this column existed has it as NULL after
183
+ # ensure_columns' ALTER (the migration is additive and never
184
+ # backfills), and json.loads(None) raises.
185
+ "pass_threads": json.loads(record.pass_threads_json or "[]"),
186
+ # Same NULL guard as pass_threads above: ensure_columns'
187
+ # ALTER never backfills, so a row written before this column
188
+ # existed reads as NULL, and json.loads(None) raises.
189
+ "candidates": json.loads(record.candidates_json or "[]"),
190
+ "final_report": record.final_report,
191
+ }
192
+
193
+
194
+ # ---------------------------------------------------------------------------
195
+ # ---- One-time classification of pre-§27 threads ---------------------------
196
+ # ---------------------------------------------------------------------------
197
+
198
+ def classify_legacy_pass_threads(connection) -> int:
199
+ """Label pass threads that predate ConversationThread.kind. Returns the
200
+ number of rows relabelled.
201
+
202
+ WHY THIS IS NOT PART OF ensure_columns(). That function is additive
203
+ only and explicitly never backfills (#21 M7) -- so every row already on
204
+ disk takes `kind='chat'` from the column default, and the hundreds of
205
+ pass threads in an existing database would still fill the picker #27
206
+ exists to clear. Widening ensure_columns' contract to backfill would
207
+ put data policy inside a schema migrator; if that function ever grows a
208
+ DROP or an UPDATE, the project has outgrown it.
209
+
210
+ WHY IT LIVES HERE rather than in storage.py. The signal is a pipeline
211
+ fact -- which threads belong to which run -- and storage.py owns schema
212
+ and CRUD without knowing what a run is. The dependency runs this way
213
+ round only; storage.py does not import this module.
214
+
215
+ TWO TIERS, EXACT FIRST (#82). Since T2, every run records the exact
216
+ thread ids of its passes in `pass_threads_json`. That record is
217
+ authoritative over any clock comparison -- so:
218
+
219
+ * TIER 1 (exact): every thread id named in a populated
220
+ `pass_threads_json` is a pass thread, whatever its timestamp says
221
+ and whether or not the run ever finished. Membership was recorded
222
+ at run time; there is nothing to infer.
223
+ * TIER 2 (window): only runs whose `pass_threads_json` is absent,
224
+ NULL, empty, or unusable fall back to the time-window heuristic --
225
+ "created between started_at and finished_at". That set is the
226
+ pre-T2 population, WHICH CANNOT GROW: a run recorded after T2
227
+ always names its own passes, so its window is never consulted and
228
+ a chat thread another session creates during it is safe (#82's
229
+ headline hazard).
230
+
231
+ CORRUPTED JSON FALLS BACK, LOUDLY. An unparseable `pass_threads_json`
232
+ puts THAT run in the window tier rather than crashing the launch --
233
+ the under-classification direction, which this docstring already
234
+ prefers -- and warns naming the run, because silently ignoring data
235
+ the writer did write is how the corruption spreads unnoticed.
236
+
237
+ A RUN WITH finished_at IS NULL GETS ONLY TIER 1. The window needs a
238
+ closing bound: #22's abandoned-generator case deliberately leaves
239
+ status='running', so an open window would swallow every thread created
240
+ since -- including real conversations. Tier 1 needs no bound (exact
241
+ ids), so an abandoned run's recorded passes ARE relabelled.
242
+ Under-classifying leaves a straggler in the picker; over-classifying
243
+ HIDES A REAL CONVERSATION, and only one of those is recoverable by the
244
+ user.
245
+
246
+ IDEMPOTENT BY CONSTRUCTION. It only ever moves rows still 'chat', so
247
+ running it at every launch is a no-op after the first -- and a database
248
+ last opened by an older build still gets classified, both for its
249
+ pre-T2 rows (window tier) and its post-T2 rows (whose ids were recorded
250
+ all along).
251
+
252
+ Raw SQL over a DBAPI connection (`engine.raw_connection()` in
253
+ production, stdlib `sqlite3.connect(...)` in tests), the same seam
254
+ database.ensure_columns() uses and for the same reason: the suite's
255
+ fake `sqlmodel` cannot execute DDL, and a seam that only works in
256
+ production is not a seam. Does NOT close the connection -- the caller
257
+ owns its lifetime. Computation happens in Python (fetch runs once,
258
+ fetch chat rows once, one bounded UPDATE per chunk) so the two tiers'
259
+ conditions are ordinary code a mutation can target, not a correlated
260
+ subquery whose bounds the suite could only probe end-to-end.
261
+
262
+ Threads a legacy compactor or subagent created are relabelled
263
+ 'research_pass' here if they happen to fall inside a PRE-T2 run's
264
+ window, and left as 'chat' otherwise. Both acceptable: the first is
265
+ hidden either way, and the second is under-classification.
266
+ """
267
+ cursor = connection.cursor()
268
+ cursor.execute("PRAGMA table_info(conversationthread)")
269
+ if not any(row[1] == "kind" for row in cursor.fetchall()):
270
+ # No column yet, so nothing to classify. create_all/ensure_columns
271
+ # run before this in production; a caller that skipped them gets a
272
+ # no-op rather than "no such column".
273
+ return 0
274
+ cursor.execute("PRAGMA table_info(pipelinerunrecord)")
275
+ run_columns = {row[1] for row in cursor.fetchall()}
276
+ if not run_columns:
277
+ # A database that has never run research. Nothing to classify, and
278
+ # neither query below would find its table.
279
+ return 0
280
+ has_recorded = "pass_threads_json" in run_columns
281
+
282
+ # ---- Fetch the runs once, and sort each into a tier -------------------
283
+ cursor.execute(
284
+ "SELECT id, started_at, finished_at"
285
+ + (", pass_threads_json" if has_recorded else "")
286
+ + " FROM pipelinerunrecord")
287
+ run_rows = cursor.fetchall()
288
+
289
+ exact_ids = set() # tier 1: named in some run's pass_threads_json
290
+ window_runs = [] # tier 2: (started_at, finished_at) pairs
291
+ for row in run_rows:
292
+ run_id, started_at, finished_at = row[0], row[1], row[2]
293
+ raw = row[3] if has_recorded else None
294
+ usable, corrupted = False, False
295
+ if raw:
296
+ try:
297
+ entries = json.loads(raw)
298
+ except ValueError:
299
+ corrupted = True
300
+ else:
301
+ if not isinstance(entries, list):
302
+ corrupted = True
303
+ else:
304
+ # T2's shape: [{"pass": ..., "thread_id": ...}]. Entries
305
+ # without a usable thread_id contribute nothing either
306
+ # way; an entry list that is merely EMPTY is the
307
+ # legitimate pre-T2 shape, i.e. window tier.
308
+ for entry in entries:
309
+ if (isinstance(entry, dict)
310
+ and isinstance(entry.get("thread_id"), str)):
311
+ exact_ids.add(entry["thread_id"])
312
+ usable = True
313
+ # raw None/"" (or a missing column) falls through with usable=False:
314
+ # the legitimate pre-T2 shape, i.e. window tier -- not corruption.
315
+ if corrupted:
316
+ # #82, owner decision B3: fall back for THIS run, and say so --
317
+ # data the writer did write must not vanish without a trace.
318
+ logger.warning(
319
+ "Classify pass threads: run %s has unparseable "
320
+ "pass_threads_json (%r); using its time window instead.",
321
+ run_id, raw[:80] if isinstance(raw, str) else raw)
322
+ if not usable:
323
+ # Window tier -- but only a CLOSED window. The intent guard
324
+ # against a future edit "kindly" treating NULL finished_at as
325
+ # open-ended lives here, in ordinary code, where removing it
326
+ # turns a test red (it never could under the old correlated
327
+ # subquery: SQL's x <= NULL was already not true).
328
+ if started_at is not None and finished_at is not None:
329
+ window_runs.append((started_at, finished_at))
330
+
331
+ # ---- Fetch the still-chat rows once ------------------------------------
332
+ cursor.execute(
333
+ "SELECT id, created_at FROM conversationthread WHERE kind = ?",
334
+ (THREAD_KIND_CHAT,))
335
+ chat_rows = cursor.fetchall()
336
+
337
+ to_relabel = []
338
+ for thread_id, created_at in chat_rows:
339
+ if thread_id in exact_ids:
340
+ to_relabel.append(thread_id)
341
+ elif created_at is not None and any(
342
+ start <= created_at <= finish for start, finish in window_runs):
343
+ to_relabel.append(thread_id)
344
+
345
+ relabelled = 0
346
+ # Chunked: SQLite's host-parameter ceiling is finite, and a long-lived
347
+ # database can carry thousands of pass threads.
348
+ for offset in range(0, len(to_relabel), 500):
349
+ chunk = to_relabel[offset:offset + 500]
350
+ placeholders = ",".join("?" * len(chunk))
351
+ cursor.execute(
352
+ f"UPDATE conversationthread SET kind = ? "
353
+ f" WHERE kind = ? AND id IN ({placeholders})",
354
+ (THREAD_KIND_RESEARCH_PASS, THREAD_KIND_CHAT, *chunk))
355
+ relabelled += cursor.rowcount if cursor.rowcount > 0 else 0
356
+ connection.commit()
357
+ if relabelled:
358
+ logger.info(
359
+ "Classified %d pre-\u00a727 thread(s) as research passes; they no "
360
+ "longer appear in the thread picker.", relabelled)
361
+ return relabelled