@vectorize-io/hindsight-coding-agents 0.8.6 → 0.8.7

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 (62) hide show
  1. package/README.md +23 -18
  2. package/dist/antigravity-hook.js +130 -135
  3. package/dist/antigravity-statusline.js +1 -1
  4. package/dist/antigravity-stop-hook.js +113 -114
  5. package/dist/claude-hook.js +130 -135
  6. package/dist/claude-sessionstart-hook.js +127 -128
  7. package/dist/claude-stop-hook.js +113 -114
  8. package/dist/cline.js +128 -124
  9. package/dist/codebuddy-hook.js +130 -135
  10. package/dist/codebuddy-sessionstart-hook.js +127 -128
  11. package/dist/codebuddy-stop-hook.js +113 -114
  12. package/dist/codex-hook.js +130 -135
  13. package/dist/codex-sessionstart-hook.js +127 -128
  14. package/dist/codex-stop-hook.js +113 -114
  15. package/dist/copilot-hook.js +130 -135
  16. package/dist/copilot-sessionstart-hook.js +127 -128
  17. package/dist/copilot-stop-hook.js +113 -114
  18. package/dist/cursor-hook.js +130 -135
  19. package/dist/cursor-sessionstart-hook.js +127 -128
  20. package/dist/cursor-stop-hook.js +113 -114
  21. package/dist/daemon-start.js +1 -1
  22. package/dist/dcode-hook.js +130 -135
  23. package/dist/dcode-sessionstart-hook.js +127 -128
  24. package/dist/dcode-stop-hook.js +113 -114
  25. package/dist/deepen.js +80 -79
  26. package/dist/devin-hook.js +130 -135
  27. package/dist/devin-sessionstart-hook.js +127 -128
  28. package/dist/devin-stop-hook.js +113 -114
  29. package/dist/droid-hook.js +130 -135
  30. package/dist/droid-sessionstart-hook.js +127 -128
  31. package/dist/droid-stop-hook.js +113 -114
  32. package/dist/dsh.js +157 -161
  33. package/dist/grok-hook.js +130 -135
  34. package/dist/grok-sessionstart-hook.js +127 -128
  35. package/dist/grok-stop-hook.js +113 -114
  36. package/dist/hindsight-seed.js +1 -1
  37. package/dist/index.js +158 -162
  38. package/dist/installer.js +30 -28
  39. package/dist/kilo.js +158 -162
  40. package/dist/kimi-hook.js +130 -135
  41. package/dist/kimi-sessionstart-hook.js +127 -128
  42. package/dist/kimi-stop-hook.js +113 -114
  43. package/dist/mcp-server.js +94 -93
  44. package/dist/opencode2.js +158 -162
  45. package/dist/pi.js +157 -161
  46. package/dist/prime-agent.js +157 -161
  47. package/dist/qwen-hook.js +130 -135
  48. package/dist/qwen-sessionstart-hook.js +127 -128
  49. package/dist/qwen-stop-hook.js +113 -114
  50. package/dist/status.js +80 -79
  51. package/dist/traecode-hook.js +130 -135
  52. package/dist/traecode-sessionstart-hook.js +127 -128
  53. package/dist/traecode-stop-hook.js +113 -114
  54. package/dist/workbuddy-hook.js +130 -135
  55. package/dist/workbuddy-sessionstart-hook.js +127 -128
  56. package/dist/workbuddy-stop-hook.js +113 -114
  57. package/dist/zcode-hook.js +130 -135
  58. package/dist/zcode-sessionstart-hook.js +127 -128
  59. package/dist/zcode-stop-hook.js +113 -114
  60. package/package.json +2 -1
  61. package/plugin.json +1 -1
  62. package/skill/SKILL.md +23 -23
package/README.md CHANGED
@@ -9,9 +9,14 @@ bank in the background as you work.
9
9
  The premise: most of a real fix is derivable from the code, but the _last mile_ often hinges on a
10
10
  project-specific decision that isn't in the code at all — a rounding rule, a retry allowlist, a
11
11
  tie-break policy. Those decisions live in git history and past conversations. This package puts them
12
- in front of the agent at the moment it starts working, and keeps a curated set of **knowledge pages**
12
+ in front of the agent at the moment it starts working, and keeps a curated set of **pages**
13
13
  (architecture, conventions, in-flight initiatives) that future sessions start from.
14
14
 
15
+ The agent reads those pages through the `hindsight_search_pages`, `hindsight_list_pages` and
16
+ `hindsight_read_page` tools. Before the rename they were `hindsight_search_knowledge_pages`,
17
+ `hindsight_list_knowledge_pages` and `hindsight_read_knowledge_page`: a permission rule in your
18
+ agent's own settings that names an old tool needs the new name.
19
+
15
20
  <!-- figure: coding-agents -->
16
21
 
17
22
  [View Changelog →](https://hindsight.vectorize.io/changelog/integrations/coding-agents)
@@ -360,7 +365,7 @@ resolved per session workspace rather than once per process; and dsh has no plug
360
365
  channel, so the seed line goes to the plugin log rather than the UI. A **forked** session saves only
361
366
  the turns it adds after the fork — the part it inherited is already in the parent's conversation.
362
367
  Everything model-facing —
363
- recalled memory, the knowledge preamble, the `hindsight_*` tools — is unaffected. If you prefer the
368
+ recalled memory, the pages preamble, the `hindsight_*` tools — is unaffected. If you prefer the
364
369
  published-package route, `dsh plugin --profile web add @vectorize-io/hindsight-coding-agents` works
365
370
  too: the package ships the profile patch layer, so nothing else needs editing. Either route gets the
366
371
  companion skill — a plugin wired by the host's own plugin manager installs it itself on the first
@@ -595,7 +600,7 @@ what reads it:
595
600
  authentication or rotating the key is picked up on the next call with nothing to restart —
596
601
  otherwise a rotation would leave a long-running agent failing every memory call until it was
597
602
  restarted. Everything else follows the table: `apiUrl`, `disabled`, bank routing, `gitIngest`, and
598
- the survey and knowledge-page settings.
603
+ the survey and page settings.
599
604
 
600
605
  `hindsight_diagnose` reports both sides of that gap — what the file says now, and what the running
601
606
  client is actually using.
@@ -663,20 +668,20 @@ hook by Codex...), so one shared config serves several agents side by side:
663
668
  | `defaultBankConfig` | — | bank-config fields of your own for the banks the plugin shapes, e.g. `{"enable_observations": false, "enable_auto_consolidation": false, "mental_model_min_refresh_interval_seconds": 21600}`. Keys are the bank-config API's own field names. Written under the same additive rule as the rest: only where the bank does not define the field, so a bank the plugin creates is born with these instead of the server's defaults and a value you set in the control plane is never overwritten. Wins over the template on a key both name (`enable_observations`); `retain_strategies`, `entity_labels` and `retain_extraction_mode` are refused, since the plugin governs them itself. Ignored with `manageBankConfig: false` — see **A bank you shape yourself** below. File-only, like `recallOptions`; in a `banks.<id>` section it replaces the global map rather than merging into it |
664
669
  | `observationScopes` | `"shared"` | how consolidation groups observations: `"shared"` (default) = ONE global scope per bank, so every agent on a repo builds one set of beliefs; also `"combined"` (the server default), `"per_tag"`, `"all_combinations"`, `[["t"]]`; `"per_source"` adds a scope per `source:` kind alongside the global one, so commit knowledge and conversation knowledge consolidate apart |
665
670
  | `disabled` | `false` | hard off-switch (inert plugin/hook — a no-memory baseline) |
666
- | `reflectTimeoutMs` | `20000` | **automatic** session-reflect timeout; on hook harnesses the installer registers a 30s prompt-hook timeout, so going above ~20s also means raising that hook's `timeout` in the host's config, or the host kills the hook mid-reflect; on timeout or a 5xx the hook falls back to knowledge-page search, then to a raw recall of observations (recorded) |
671
+ | `reflectTimeoutMs` | `20000` | **automatic** session-reflect timeout; on hook harnesses the installer registers a 30s prompt-hook timeout, so going above ~20s also means raising that hook's `timeout` in the host's config, or the host kills the hook mid-reflect; on timeout or a 5xx the hook falls back to page search, then to a raw recall of observations (recorded) |
667
672
  | `reflectToolTimeoutMs` | `330000` | timeout for the agent-invoked `hindsight_reflect` tool — a call the agent waits on, whose high-budget synthesis on a populated bank runs for minutes. Defaults above the server's own reflect wall timeout (`HINDSIGHT_API_REFLECT_WALL_TIMEOUT`, 300s) so the server decides when to give up. Unset, it inherits an explicitly raised `reflectTimeoutMs`, but a short one never lowers it |
668
673
  | `injectTimeoutMs` | `7000` | retrieval timeout in milliseconds for `autoInject: "pages"` and `autoInject: "recall"`; also the **shared** budget for page search then recall after a reflect timeout/5xx (recall gets only the time left after page search). Set e.g. `2000` to limit turn latency or `15000` for a slower self-hosted bank. Environment fallback: `HINDSIGHT_INJECT_TIMEOUT_MS`. On hook harnesses, injection must fit the host's prompt-hook timeout (30s by default); for reflect with fallback, allow for **`reflectTimeoutMs` + `injectTimeoutMs`**, and raise the host hook's `timeout` when needed |
669
674
  | `reflectBudget` | `"high"` | reflect budget for the `hindsight_reflect` tool: `"low"`, `"mid"` or `"high"`. Drop it on a large bank where high-budget synthesis exceeds the server's wall timeout. The automatic session-start reflect always uses `"low"` to fit its hook window and is unaffected |
670
- | `autoInject` | `"reflect"` | what to inject once, on the session's **first prompt**: `"reflect"` = a low-budget reflect synthesis (on timeout/5xx it falls back to page search, then recall); `"pages"` = the knowledge pages matching the prompt by search; `"recall"` = the bank's memories recalled for the prompt (what it asks for is `recallOptions`, observations by default); `"none"` = nothing — the agent searches knowledge pages first and reflects only when they are too shallow. `"pages"` and `"recall"` are retrieval only (no LLM), their time budget is `injectTimeoutMs`, and when one finds nothing for the prompt it tries again on the next one, up to 5 prompts |
675
+ | `autoInject` | `"reflect"` | what to inject once, on the session's **first prompt**: `"reflect"` = a low-budget reflect synthesis (on timeout/5xx it falls back to page search, then recall); `"pages"` = the pages matching the prompt by search; `"recall"` = the bank's memories recalled for the prompt (what it asks for is `recallOptions`, observations by default); `"none"` = nothing — the agent searches pages first and reflects only when they are too shallow. `"pages"` and `"recall"` are retrieval only (no LLM), their time budget is `injectTimeoutMs`, and when one finds nothing for the prompt it tries again on the next one, up to 5 prompts |
671
676
  | `autoReflect` | `true` | **deprecated** — use `autoInject`. Still honoured (`false` = `autoInject: "none"`), ignored when `autoInject` is set, and logs a deprecation warning |
672
- | `pageSearchLimit` | `3` | knowledge pages ONE search returns. Applies to every search of the bank — the `autoInject: "pages"` injection, the reflect fallback, and the agent's own `hindsight_search_knowledge_pages` tool — because the limit lives on the client they all share |
677
+ | `pageSearchLimit` | `3` | pages ONE search returns. Applies to every search of the bank — the `autoInject: "pages"` injection, the reflect fallback, and the agent's own `hindsight_search_pages` tool — because the limit lives on the client they all share |
673
678
  | `recallOptions` | see description | overrides merged key-by-key into the body of every recall — the `autoInject: "recall"` source and the reflect fallback. Keys are the API's own recall parameters, passed straight through, so anything recall accepts is settable without a new option here; `query` is the one field it cannot replace. The default body is `{"types": ["observation"], "budget": "low", "max_tokens": 2000, "include": {"entities": null}}`, and what you set is merged over it one key at a time — `{"max_tokens": 4000}` changes the budget and leaves the rest alone. Observations are the consolidated layer, so they answer best per token, but a bank with **consolidation disabled never grows any** and the default recall comes back empty on it: set `{"types": ["world", "experience"]}` there, or `{"types": null}` for every type. File-only, like `retainMetadata` — an object does not flatten into an env var |
674
- | `toolGuideExtra` | — | your own guidance on how the agent should treat memory, added **after** the built-in tool guide (at session start and on every refresh) and after the crediting note returned with `hindsight_search_knowledge_pages` results. It adds to the built-in text, never replaces it. Set it per harness or per bank like any other key, e.g. `"Memory is a past record, not current state: verify any claim against the code before acting on it."` |
675
- | `pageRefreshEveryTurns` | `1` | refetch the knowledge pages and re-inject the page roster + tool guide every N user turns (`0` also means 1) |
676
- | `pageTriggerType` | `"cron"` | when NEW knowledge pages refresh, i.e. what keeping them current costs — `"cron"` (default) on `pageTriggerCron` only and only when actually stale, `"auto-refresh"` after every consolidation that produced new material, `"manual"` never on their own. Auto-refresh is the most current and by far the most expensive: one synthesis per page per consolidation. Maps to the page's `trigger.refresh_cron`, or `trigger.refresh_after_consolidation` in the Hindsight API (`true` for auto-refresh, `false` for manual) |
679
+ | `toolGuideExtra` | — | your own guidance on how the agent should treat memory, added **after** the built-in tool guide (at session start and on every refresh) and after the crediting note returned with `hindsight_search_pages` results. It adds to the built-in text, never replaces it. Set it per harness or per bank like any other key, e.g. `"Memory is a past record, not current state: verify any claim against the code before acting on it."` |
680
+ | `pageRefreshEveryTurns` | `1` | refetch the pages and re-inject the page roster + tool guide every N user turns (`0` also means 1) |
681
+ | `pageTriggerType` | `"cron"` | when NEW pages refresh, i.e. what keeping them current costs — `"cron"` (default) on `pageTriggerCron` only and only when actually stale, `"auto-refresh"` after every consolidation that produced new material, `"manual"` never on their own. Auto-refresh is the most current and by far the most expensive: one synthesis per page per consolidation. Maps to the page's `trigger.refresh_cron`, or `trigger.refresh_after_consolidation` in the Hindsight API (`true` for auto-refresh, `false` for manual) |
677
682
  | `pageTriggerCron` | `"H * * * *"` | schedule for `pageTriggerType: "cron"` — UTC, standard 5-field cron, e.g. `"0 3 * * *"`. The default is hourly, each page on its own hashed minute. Sets the page's `trigger.refresh_cron`, which the API treats as mutually exclusive with `refresh_after_consolidation`; a scheduled refresh is skipped when nothing changed. Write a field as `H` to give each page its own value there — see **Spreading refreshes with `H`** below |
678
- | `pages` | every page | per-page configuration for the seeded knowledge pages, keyed by page name (case-insensitive): `false` skips a page entirely, `{"source_query": "..."}` seeds it with your question instead of the built-in one. Omitted, all five pages are seeded with their built-in queries. This is the supported way to own a page's wording — the plugin re-syncs a page whose live query differs from the one it is configured to have, so a query edited through the API or the control plane is replaced on the next session. A skipped page is **not deleted**: one already seeded keeps its content and stops being re-synced. The scoping clause is appended to your query too, so a reworded page cannot start reporting a dependency's decisions as this project's. File-only, like `recallOptions`; in a `banks.<id>` section it replaces the global map rather than merging into it |
679
- | `customPages` | — | knowledge pages of your own, seeded alongside the five above and keyed by the name they get: `{"Security posture": {"source_query": "...", "tags": ["knowledge:decision"]}}`. `source_query` is required; `tags` picks which facts feed the page and is optional — omitted, the page draws on everything the bank holds. A separate setting from `pages` on purpose, so that an unknown name there stays a typo warning rather than quietly creating a page. File-only, and replaced (not merged) by a `banks.<id>` section |
683
+ | `pages` | every page | per-page configuration for the seeded pages, keyed by page name (case-insensitive): `false` skips a page entirely, `{"source_query": "..."}` seeds it with your question instead of the built-in one. Omitted, all five pages are seeded with their built-in queries. This is the supported way to own a page's wording — the plugin re-syncs a page whose live query differs from the one it is configured to have, so a query edited through the API or the control plane is replaced on the next session. A skipped page is **not deleted**: one already seeded keeps its content and stops being re-synced. The scoping clause is appended to your query too, so a reworded page cannot start reporting a dependency's decisions as this project's. File-only, like `recallOptions`; in a `banks.<id>` section it replaces the global map rather than merging into it |
684
+ | `customPages` | — | pages of your own, seeded alongside the five above and keyed by the name they get: `{"Security posture": {"source_query": "...", "tags": ["knowledge:decision"]}}`. `source_query` is required; `tags` picks which facts feed the page and is optional — omitted, the page draws on everything the bank holds. A separate setting from `pages` on purpose, so that an unknown name there stays a typo warning rather than quietly creating a page. File-only, and replaced (not merged) by a `banks.<id>` section |
680
685
  | `autoSeed` | `true` | SessionStart: auto-seed a cold repo's bank from git history |
681
686
  | `seedLimit` | `300` | git-log document size: a full write keeps the newest commits, between half of this and this many (the window moves in steps of half), and logs how many older ones it drops |
682
687
  | `codebaseSurvey` | `true` | SessionStart: headless survey of a cold repo's structure, run only under the current harness's own CLI (Claude Code, Codex, Antigravity, opencode v1); other agents get no survey |
@@ -698,7 +703,7 @@ LLM synthesis per page per consolidation, on a repo that consolidates all day. S
698
703
  `pageTriggerType: "auto-refresh"` to go back to refreshing on every consolidation.
699
704
 
700
705
  `pageTriggerType`/`pageTriggerCron` decide only **when** a page refreshes. **How** it refreshes
701
- belongs to the server: Hindsight creates a knowledge page with a delta refresh (each pass edits the
706
+ belongs to the server: Hindsight creates a page with a delta refresh (each pass edits the
702
707
  page instead of rebuilding it) that doesn't reflect over sibling pages, and these settings merge
703
708
  over those defaults rather than replacing them.
704
709
 
@@ -736,7 +741,7 @@ file is the source of truth for these pages: to give one a different schedule, c
736
741
  page. Only the fields this plugin states are touched — a page's `mode`, its excluded siblings and
737
742
  its minimum refresh interval are left exactly as they are.
738
743
 
739
- ### Customize Knowledge Pages — `pages`
744
+ ### Customize Pages — `pages`
740
745
 
741
746
  Every repo gets the same five pages. They are a **taxonomy**, not a summary of your source: each one
742
747
  is synthesized from what the bank ingested — commit history and past conversations — and each is
@@ -857,7 +862,7 @@ it: the seed pass only visits the pages it is configured to own.
857
862
 
858
863
  Pointed at a bank, this plugin gives it the shape its ingestion needs: retain strategies for the
859
864
  kinds of document it writes (`git`, `gitlog`, `conversation`, `document`, `survey`), a `knowledge`
860
- entity-label group that routes facts to the knowledge pages, and — on a bank that has no missions of
865
+ entity-label group that routes facts to the pages, and — on a bank that has no missions of
861
866
  its own — the coding missions.
862
867
 
863
868
  **It only ever adds what is missing.** A strategy you defined, an edit you made to one of the
@@ -906,7 +911,7 @@ you configure yourself. That bank should then define the five strategies above i
906
911
  miss is **silent**: the server does not reject a retain naming a strategy the bank lacks, it logs a
907
912
  warning and extracts with the bank's own configuration — so a commit diff, a session transcript and
908
913
  a survey marker would all get the same generic treatment instead of the extraction each needs.
909
- Knowledge pages are seeded either way; `pageTriggerType` governs what they cost.
914
+ Pages are seeded either way; `pageTriggerType` governs what they cost.
910
915
 
911
916
  Like every field here it can be set per bank, which is usually where it belongs:
912
917
 
@@ -1097,7 +1102,7 @@ scope fills with beliefs built from chat transcripts. Only a per-document decisi
1097
1102
 
1098
1103
  It costs one extra consolidation pass per document, and it reads only `source:`, so a volatile
1099
1104
  provenance tag never becomes a scope. The global scope is still written first and unchanged, so the
1100
- untagged observations knowledge pages read are unaffected.
1105
+ untagged observations pages read are unaffected.
1101
1106
 
1102
1107
  `"per_tag"` and `"all_combinations"` split further still, and an explicit `[["project:demo"], …]`
1103
1108
  declares the scopes literally. `HINDSIGHT_OBSERVATION_SCOPES` sets the scalar modes; a scope list is
@@ -1111,7 +1116,7 @@ they stay where they were built, and new work accrues under the new setting.
1111
1116
  There is no user-facing ingest command — the deepen engine (`dist/deepen.js`) is spawned by every
1112
1117
  session start and does only the missing work: bank configuration, conversation import (dedup by
1113
1118
  document id), the one-time gitlog seed, the next per-commit diff batch (newest first, bounded per
1114
- run), then knowledge pages once extraction has drained. Harnesses that need deterministic ingestion
1119
+ run), then pages once extraction has drained. Harnesses that need deterministic ingestion
1115
1120
  (benchmarks, e2e suites) run the same engine directly and poll `dist/status.js` until
1116
1121
  `"synced": true` — the exact readiness contract the `hindsight_sync_status` agent tool reports.
1117
1122
 
@@ -1177,7 +1182,7 @@ as a JSON line to `~/.hindsight/coding-agents-logs/diag.jsonl` (override with
1177
1182
 
1178
1183
  `reflect_failed` / `pages_failed` record the error; if you're comparing memory-on vs memory-off,
1179
1184
  check this file — a run whose reflects failed is a no-memory run. When the failure was a timeout or
1180
- a 5xx, the hook falls back to knowledge-page search and, if no page matches, to a raw recall of the
1185
+ a 5xx, the hook falls back to page search and, if no page matches, to a raw recall of the
1181
1186
  bank's memories: `reflect_fallback_pages` / `reflect_fallback_observations` record what each
1182
1187
  step returned (`*_failed` when it errored). Seed starts are logged as
1183
1188
  `seed_started`.