Replace the elicitation path with a plain Petrinaut-to-Flue chat. by lunelson · Pull Request #9382 · hashintel/hash

lunelson · 2026-08-27T15:55:35Z

🌟 What is the purpose of this PR?

🔗 Related links

...

🚫 Blocked by

...

🔍 What does this change?

...

Pre-Merge Checklist 🚀

🚢 Has this modified a publishable library?

This PR:

does not modify any publishable blocks or libraries, or modifications do not need publishing
modifies an npm-publishable library and I have added a changeset file(s)
modifies a Cargo-publishable library and I have amended the version
modifies a Cargo-publishable library, but it is not yet ready to publish
modifies a block that will need publishing via GitHub action once merged
I am unsure / need advice

📜 Does this require a change to the docs?

The changes in this PR:

are internal and do not require a docs change
are in a state where docs changes are not yet required but will be
- this is tracked in: Insert Link Here
require changes to docs which are made as part of this PR
require changes to docs which are not made in this PR
- Provide more detail here
I am unsure / need advice

🕸️ Does this require a change to the Turbo Graph?

The changes in this PR:

do not affect the execution graph
affected the execution graph, and the turbo.json's have been updated to reflect this
I am unsure / need advice

⚠️ Known issues

🐾 Next steps

🛡 What tests cover this?

❓ How to test this?

Checkout the branch / view the deployment
Try X
Confirm that Y

📹 Demo

vercel · 2026-08-27T15:55:46Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
hash	Ready	Preview	Aug 27, 2026 5:56pm
petrinaut	Ready	Preview	Aug 27, 2026 5:56pm
petrinaut-docs	Ready	Preview	Aug 27, 2026 5:56pm

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
hashdotdesign-tokens	Ignored	Preview	Aug 27, 2026 5:56pm

lunelson · 2026-08-27T15:56:08Z

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

This stack of pull requests is managed by Graphite. Learn more about stacking.

cursor · 2026-08-27T15:58:00Z

PR Summary

Medium Risk
Rewrites the production chat door, Flue route authorization, and persistence model while removing elicitation; new ownership checks are covered by tests but the surface area is still broad.

Overview
Replaces the SDCPN/Gherkin elicitation harness on the Petrinaut panel with a single plain Flue ChatAgent (ping server tool + browser readPetrinautDoc), wired through /api/chat with AI SDK streaming, GET history hydration, and client-tool resume via Flue signals—no brunch_ask, capture store, or target-document persistence on this path.

Conversation identity is now principal + panel conversation id → SHA-256 Flue instance id, with agentOwnershipGuard on /agents/chat/:id so the path id is not a bearer token. New flue-ui-stream / flue-transcript bridge Flue chunks and history() to AI SDK messages; a transcript CLI prints the same log.

Dependencies and tests drop @hashintel/brunch-agent core, binding-flue, and plugins; the app depends on @hashintel/petrinaut-core for doc tool catalogs. Large elicitation suites (baseline harness, walking skeleton, ask integration) are removed in favor of chat-path integration tests.

Petrinaut local demo pins a stable conversation id per net, hydrates from GET /api/chat, removes the Brunch ask widget and sweep stream decoration, and documents the voice dock as sharing the same POST contract. Lint/docs/README and dev:brunch now build Petrinaut libs first; Brunch MISSION.md advances to Mission 2 (mechanical capture sweep) with MISSION.next.md scratchpad guidance in AGENTS.md.

^{Reviewed by Cursor Bugbot for commit 00e16a5. Bugbot is set up for automated code reviews on this repo. Configure here.}

cursor · 2026-08-27T16:03:37Z

+        ? flueHistory.messages
+        : currentNetId
+          ? aiMessagesByNetId[currentNetId]
+          : undefined,


Flue history never hydrates the panel

High Severity

GET /api/chat history is applied by swapping aiAssistant.messages after mount, but Petrinaut's useChat only reads that prop as initial state. The panel therefore keeps the first paint's localStorage (or empty) transcript and never reconciles to Flue, so reload and cross-client turns do not reconstruct the conversation from server history.

Additional Locations (1)

apps/petrinaut-website/src/main/app/local-storage-demo/use-flue-chat-history.ts#L18-L56

^{Reviewed by Cursor Bugbot for commit 7d400c4. Configure here.}

cursor · 2026-08-27T16:03:37Z

        });
      },
    }),
-    [aiMessagesByNetId, currentNetId, setAiMessagesByNetId],


Clear chat keeps the Flue conversation

High Severity

Clearing the assistant only deletes the localStorage cache. The per-net conversationId is unchanged, so the next send still targets the same Flue instance. The model continues the old thread even though the panel looks empty, and a later successful history load would restore the wiped transcript.

Additional Locations (2)

apps/petrinaut-website/src/main/app/local-storage-demo/brunch-conversation-id.ts#L7-L30

apps/petrinaut-website/src/main/app/local-storage-demo/local-storage-demo-app.tsx#L264-L267

^{Reviewed by Cursor Bugbot for commit 7d400c4. Configure here.}

cursor · 2026-08-27T16:03:37Z

+    return { messages: [] };
+  }
+  return { messages: snapshotToUiMessages(snapshot) };
+};


History failures look like empty chats

Medium Severity

loadHistory catches every history() failure and returns HTTP 200 with messages: []. The panel treats a successful GET as authoritative Flue state, so a missing instance, timeout, or Flue error is indistinguishable from a brand-new conversation and can wipe the visible transcript.

Additional Locations (1)

apps/petrinaut-website/src/main/app/local-storage-demo/use-flue-chat-history.ts#L29-L44

^{Reviewed by Cursor Bugbot for commit 7d400c4. Configure here.}

codspeed-hq · 2026-08-27T16:06:56Z

Merging this PR will not alter performance

✅ 98 untouched benchmarks

_{Comparing ln/fe-1522-mission-1 (00e16a5) with main (3b2d8e3)¹}

No successful run was found on ln/fe-1521-machete-protocol (b06ac00) during the generation of this report, so main (3b2d8e3) was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

github-actions · 2026-08-27T17:12:06Z

Benchmark results

@rust/hash-graph-benches – Integrations

policy_resolution_large

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 2002	$$26.9 \mathrm{ms} \pm 207 \mathrm{μs}\left({\color{gray}-3.041 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$3.45 \mathrm{ms} \pm 27.1 \mathrm{μs}\left({\color{gray}-1.786 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 1002	$$12.3 \mathrm{ms} \pm 116 \mathrm{μs}\left({\color{gray}-0.596 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 3314	$$42.9 \mathrm{ms} \pm 425 \mathrm{μs}\left({\color{gray}-0.841 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$14.1 \mathrm{ms} \pm 134 \mathrm{μs}\left({\color{gray}-4.815 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 1527	$$23.7 \mathrm{ms} \pm 253 \mathrm{μs}\left({\color{gray}-0.731 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 2078	$$28.5 \mathrm{ms} \pm 269 \mathrm{μs}\left({\color{gray}0.365 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.76 \mathrm{ms} \pm 22.6 \mathrm{μs}\left({\color{gray}-2.410 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 1033	$$13.4 \mathrm{ms} \pm 146 \mathrm{μs}\left({\color{gray}0.213 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_medium

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 102	$$3.84 \mathrm{ms} \pm 30.5 \mathrm{μs}\left({\color{gray}0.239 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$3.00 \mathrm{ms} \pm 21.6 \mathrm{μs}\left({\color{gray}-0.082 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 52	$$3.37 \mathrm{ms} \pm 25.6 \mathrm{μs}\left({\color{gray}-0.923 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 269	$$5.19 \mathrm{ms} \pm 39.7 \mathrm{μs}\left({\color{gray}-1.090 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$3.54 \mathrm{ms} \pm 23.4 \mathrm{μs}\left({\color{gray}-0.562 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 108	$$4.15 \mathrm{ms} \pm 37.6 \mathrm{μs}\left({\color{gray}-1.056 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 133	$$4.41 \mathrm{ms} \pm 28.0 \mathrm{μs}\left({\color{gray}-1.793 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$3.49 \mathrm{ms} \pm 23.0 \mathrm{μs}\left({\color{gray}0.533 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 63	$$4.09 \mathrm{ms} \pm 27.8 \mathrm{μs}\left({\color{gray}-0.703 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_none

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 2	$$2.69 \mathrm{ms} \pm 22.1 \mathrm{μs}\left({\color{gray}3.15 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$2.54 \mathrm{ms} \pm 16.8 \mathrm{μs}\left({\color{gray}0.431 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 2	$$2.66 \mathrm{ms} \pm 15.4 \mathrm{μs}\left({\color{gray}2.86 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 8	$$2.94 \mathrm{ms} \pm 17.4 \mathrm{μs}\left({\color{gray}1.99 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$2.74 \mathrm{ms} \pm 16.8 \mathrm{μs}\left({\color{gray}3.91 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 3	$$2.98 \mathrm{ms} \pm 31.9 \mathrm{μs}\left({\color{gray}4.56 \mathrm{\%}}\right) $$	Flame Graph

policy_resolution_small

Function	Value	Mean	Flame graphs
resolve_policies_for_actor	user: empty, selectivity: high, policies: 52	$$3.06 \mathrm{ms} \pm 26.6 \mathrm{μs}\left({\color{gray}0.340 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: low, policies: 1	$$2.76 \mathrm{ms} \pm 18.7 \mathrm{μs}\left({\color{gray}1.15 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: empty, selectivity: medium, policies: 26	$$3.00 \mathrm{ms} \pm 23.2 \mathrm{μs}\left({\color{gray}2.00 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: high, policies: 94	$$3.46 \mathrm{ms} \pm 25.1 \mathrm{μs}\left({\color{gray}-0.095 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: low, policies: 1	$$3.02 \mathrm{ms} \pm 28.1 \mathrm{μs}\left({\color{gray}2.21 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: seeded, selectivity: medium, policies: 27	$$3.32 \mathrm{ms} \pm 28.1 \mathrm{μs}\left({\color{gray}2.31 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: high, policies: 66	$$3.39 \mathrm{ms} \pm 31.1 \mathrm{μs}\left({\color{gray}0.451 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: low, policies: 1	$$2.96 \mathrm{ms} \pm 25.3 \mathrm{μs}\left({\color{gray}-0.028 \mathrm{\%}}\right) $$	Flame Graph
resolve_policies_for_actor	user: system, selectivity: medium, policies: 29	$$3.35 \mathrm{ms} \pm 25.8 \mathrm{μs}\left({\color{gray}1.95 \mathrm{\%}}\right) $$	Flame Graph

read_scaling_complete

Function	Value	Mean	Flame graphs
entity_by_id;one_depth	1 entities	$$41.1 \mathrm{ms} \pm 241 \mathrm{μs}\left({\color{gray}-0.752 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	10 entities	$$31.8 \mathrm{ms} \pm 236 \mathrm{μs}\left({\color{gray}-2.038 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	25 entities	$$34.2 \mathrm{ms} \pm 324 \mathrm{μs}\left({\color{gray}-0.410 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	5 entities	$$31.2 \mathrm{ms} \pm 278 \mathrm{μs}\left({\color{gray}-0.960 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;one_depth	50 entities	$$40.6 \mathrm{ms} \pm 243 \mathrm{μs}\left({\color{gray}-4.204 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	1 entities	$$48.9 \mathrm{ms} \pm 304 \mathrm{μs}\left({\color{gray}-0.017 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	10 entities	$$39.2 \mathrm{ms} \pm 271 \mathrm{μs}\left({\color{gray}0.012 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	25 entities	$$89.6 \mathrm{ms} \pm 752 \mathrm{μs}\left({\color{gray}-1.023 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	5 entities	$$32.8 \mathrm{ms} \pm 215 \mathrm{μs}\left({\color{gray}-0.452 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;two_depth	50 entities	$$305 \mathrm{ms} \pm 1.60 \mathrm{ms}\left({\color{red}9.64 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	1 entities	$$10.4 \mathrm{ms} \pm 91.9 \mathrm{μs}\left({\color{gray}0.382 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	10 entities	$$10.5 \mathrm{ms} \pm 70.8 \mathrm{μs}\left({\color{gray}-0.605 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	25 entities	$$10.4 \mathrm{ms} \pm 92.5 \mathrm{μs}\left({\color{gray}0.360 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	5 entities	$$10.4 \mathrm{ms} \pm 82.1 \mathrm{μs}\left({\color{gray}-0.822 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id;zero_depth	50 entities	$$10.4 \mathrm{ms} \pm 74.2 \mathrm{μs}\left({\color{gray}-0.084 \mathrm{\%}}\right) $$	Flame Graph

read_scaling_linkless

Function	Value	Mean	Flame graphs
entity_by_id	1 entities	$$10.4 \mathrm{ms} \pm 68.1 \mathrm{μs}\left({\color{gray}-0.349 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	10 entities	$$10.5 \mathrm{ms} \pm 78.0 \mathrm{μs}\left({\color{gray}0.083 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	100 entities	$$10.5 \mathrm{ms} \pm 67.4 \mathrm{μs}\left({\color{gray}-0.689 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	1000 entities	$$10.5 \mathrm{ms} \pm 79.6 \mathrm{μs}\left({\color{gray}-0.159 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	10000 entities	$$10.5 \mathrm{ms} \pm 66.0 \mathrm{μs}\left({\color{gray}-1.583 \mathrm{\%}}\right) $$	Flame Graph

representative_read_entity

Function	Value	Mean	Flame graphs
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/block/v/1`	$$10.8 \mathrm{ms} \pm 65.5 \mathrm{μs}\left({\color{gray}-2.871 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/book/v/1`	$$10.8 \mathrm{ms} \pm 60.0 \mathrm{μs}\left({\color{gray}-2.435 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/building/v/1`	$$10.9 \mathrm{ms} \pm 69.5 \mathrm{μs}\left({\color{gray}-0.609 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/organization/v/1`	$$10.9 \mathrm{ms} \pm 70.8 \mathrm{μs}\left({\color{gray}0.486 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/page/v/2`	$$10.7 \mathrm{ms} \pm 58.5 \mathrm{μs}\left({\color{gray}-0.668 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/person/v/1`	$$11.2 \mathrm{ms} \pm 101 \mathrm{μs}\left({\color{gray}1.03 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/playlist/v/1`	$$10.7 \mathrm{ms} \pm 60.4 \mathrm{μs}\left({\color{gray}-1.364 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/song/v/1`	$$10.7 \mathrm{ms} \pm 68.9 \mathrm{μs}\left({\color{gray}-1.785 \mathrm{\%}}\right) $$	Flame Graph
entity_by_id	entity type ID: `https://blockprotocol.org/@alice/types/entity-type/uk-address/v/1`	$$10.8 \mathrm{ms} \pm 74.0 \mathrm{μs}\left({\color{gray}-1.213 \mathrm{\%}}\right) $$	Flame Graph

representative_read_entity_type

Function	Value	Mean	Flame graphs
get_entity_type_by_id	Account ID: `bf5a9ef5-dc3b-43cf-a291-6210c0321eba`	$$8.05 \mathrm{ms} \pm 49.7 \mathrm{μs}\left({\color{gray}-0.928 \mathrm{\%}}\right) $$	Flame Graph

representative_read_multiple_entities

Function	Value	Mean	Flame graphs
entity_by_property	traversal_paths=0	0	$$56.2 \mathrm{ms} \pm 486 \mathrm{μs}\left({\color{gray}1.85 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=255	1,resolve_depths=inherit:1;values:255;properties:255;links:127;link_dests:126;type:true	$$109 \mathrm{ms} \pm 640 \mathrm{μs}\left({\color{gray}1.09 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:0;link_dests:0;type:false	$$63.5 \mathrm{ms} \pm 501 \mathrm{μs}\left({\color{gray}-0.414 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:1;link_dests:0;type:true	$$73.5 \mathrm{ms} \pm 509 \mathrm{μs}\left({\color{gray}-1.003 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:2;links:1;link_dests:0;type:true	$$82.7 \mathrm{ms} \pm 622 \mathrm{μs}\left({\color{gray}0.839 \mathrm{\%}}\right) $$
entity_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:2;properties:2;links:1;link_dests:0;type:true	$$87.7 \mathrm{ms} \pm 538 \mathrm{μs}\left({\color{gray}0.436 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=0	0	$$41.7 \mathrm{ms} \pm 310 \mathrm{μs}\left({\color{gray}-0.257 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=255	1,resolve_depths=inherit:1;values:255;properties:255;links:127;link_dests:126;type:true	$$70.9 \mathrm{ms} \pm 551 \mathrm{μs}\left({\color{gray}-0.282 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:0;link_dests:0;type:false	$$48.0 \mathrm{ms} \pm 392 \mathrm{μs}\left({\color{gray}0.065 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:0;links:1;link_dests:0;type:true	$$57.3 \mathrm{ms} \pm 400 \mathrm{μs}\left({\color{gray}0.032 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:0;properties:2;links:1;link_dests:0;type:true	$$59.7 \mathrm{ms} \pm 396 \mathrm{μs}\left({\color{gray}0.385 \mathrm{\%}}\right) $$
link_by_source_by_property	traversal_paths=2	1,resolve_depths=inherit:0;values:2;properties:2;links:1;link_dests:0;type:true	$$59.6 \mathrm{ms} \pm 464 \mathrm{μs}\left({\color{gray}0.015 \mathrm{\%}}\right) $$

scenarios

Function	Value	Mean	Flame graphs
full_test	query-limited	$$127 \mathrm{ms} \pm 694 \mathrm{μs}\left({\color{red}5.21 \mathrm{\%}}\right) $$	Flame Graph
full_test	query-unlimited	$$137 \mathrm{ms} \pm 688 \mathrm{μs}\left({\color{gray}3.13 \mathrm{\%}}\right) $$	Flame Graph
linked_queries	query-limited	$$18.3 \mathrm{ms} \pm 176 \mathrm{μs}\left({\color{lightgreen}-17.290 \mathrm{\%}}\right) $$	Flame Graph
linked_queries	query-unlimited	$$522 \mathrm{ms} \pm 1.23 \mathrm{ms}\left({\color{gray}-3.375 \mathrm{\%}}\right) $$	Flame Graph

Mission 1 needs a durable conversation through the real panel before capture lands, so /api/chat now mounts one Flue agent with a server tool, a read-only client-tool resume, and Flue history as the log. Co-authored-by: Cursor <cursoragent@cursor.com>

…ries. The agent module stays a thin hook composition; skills, subagents, and channels sit empty until they are used. Co-authored-by: Cursor <cursoragent@cursor.com>

…uth. The Flue tool now shares petrinaut-core's catalog instead of a free string, and yarn dev:brunch builds the panel's workspace libraries first. Co-authored-by: Cursor <cursoragent@cursor.com>

… them. The conversation projector now maps data-part and message-metadata, fails compilation on new Flue chunk types, and keeps the same parts in history reconstruction. Co-authored-by: Cursor <cursoragent@cursor.com>

The mounted agent URL is no longer a capability hash; later concerns move to the mission scratchpad so Mission 1 stays the only execution authority. Co-authored-by: Cursor <cursoragent@cursor.com>

Co-authored-by: Cursor <cursoragent@cursor.com>

cursor

Cursor Bugbot has reviewed your changes and found 2 potential issues.

There are 5 total unresolved issues (including 3 from previous reviews).

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 00e16a5. Configure here.}

cursor · 2026-08-27T17:46:48Z

+    return part.output;
+  }
+  return clientOutput;
+};


Pending client tools hydrate as complete

Medium Severity

When a client tool is still awaiting a browser result, resolveToolOutput falls back to the { awaiting: "client" } sentinel and toolPartFrom emits output-available. Reload hydration then looks like the tool already finished, so the panel will not re-run it or POST a client-tool-result, and that Flue turn stays blocked.

Additional Locations (1)

apps/brunch-agent/src/flue-transcript.ts#L110-L146

^{Reviewed by Cursor Bugbot for commit 00e16a5. Configure here.}

cursor · 2026-08-27T17:46:48Z

+      },
+    },
+    write,
+  );


Turns drop the idempotency key

Medium Severity

runUserTurn and runClientToolResume hash identity and dispatch the text or signal, but they never forward idempotencyKey. The previous elicitor path passed that key into agent.dispatch, and the transport still computes it. Retried POSTs can therefore run the same Flue turn twice.

^{Reviewed by Cursor Bugbot for commit 00e16a5. Configure here.}

lunelson mentioned this pull request Aug 27, 2026

Cut the agentic guidance layer to a mission charter #9381

Open

15 tasks

lunelson deployed to pull-request August 27, 2026 15:56 — with GitHub Actions Active

lunelson marked this pull request as ready for review August 27, 2026 15:57

lunelson marked this pull request as draft August 27, 2026 15:57

graphite-app Bot assigned lunelson Aug 27, 2026

lunelson marked this pull request as ready for review August 27, 2026 15:57

vercel Bot deployed to Preview – petrinaut-docs August 27, 2026 15:58 View deployment

vercel Bot deployed to Preview – petrinaut August 27, 2026 16:00 View deployment

cursor Bot reviewed Aug 27, 2026

View reviewed changes

vercel Bot deployed to Preview – hash August 27, 2026 16:09 View deployment

lunelson mentioned this pull request Aug 27, 2026

FE-1524: Prove idempotent capture sweep on the live Flue chat path #9384

Open

4 tasks

lunelson and others added 6 commits August 27, 2026 19:41

Move chat tools into src/tools and reserve unused Flue layout directo…

0cfd7c1

…ries. The agent module stays a thin hook composition; skills, subagents, and channels sit empty until they are used. Co-authored-by: Cursor <cursoragent@cursor.com>

Import Petrinaut doc names as the readPetrinautDoc input source of tr…

f5d2b78

…uth. The Flue tool now shares petrinaut-core's catalog instead of a free string, and yarn dev:brunch builds the panel's workspace libraries first. Co-authored-by: Cursor <cursoragent@cursor.com>

Enforce Flue-route ownership from principal and conversation identity.

d47428d

The mounted agent URL is no longer a capability hash; later concerns move to the mission scratchpad so Mission 1 stays the only execution authority. Co-authored-by: Cursor <cursoragent@cursor.com>

Archive Mission 1 and cut the mechanical capture-sweep mission.

00e16a5

Co-authored-by: Cursor <cursoragent@cursor.com>

lunelson force-pushed the ln/fe-1521-machete-protocol branch from 69dac29 to b06ac00 Compare August 27, 2026 17:42

lunelson force-pushed the ln/fe-1522-mission-1 branch from 7d400c4 to 00e16a5 Compare August 27, 2026 17:42

lunelson deployed to pull-request August 27, 2026 17:43 — with GitHub Actions Active

vercel Bot deployed to Preview – petrinaut-docs August 27, 2026 17:45 View deployment

vercel Bot deployed to Preview – petrinaut August 27, 2026 17:46 View deployment

cursor Bot reviewed Aug 27, 2026

View reviewed changes

vercel Bot deployed to Preview – hash August 27, 2026 17:56 View deployment

Uh oh!

Conversation

lunelson commented Aug 27, 2026

🌟 What is the purpose of this PR?

🔗 Related links

🚫 Blocked by

🔍 What does this change?

Pre-Merge Checklist 🚀

🚢 Has this modified a publishable library?

📜 Does this require a change to the docs?

🕸️ Does this require a change to the Turbo Graph?

⚠️ Known issues

🐾 Next steps

🛡 What tests cover this?

❓ How to test this?

📹 Demo

Uh oh!

vercel Bot commented Aug 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lunelson commented Aug 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cursor Bot commented Aug 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Summary

Uh oh!

cursor Bot Aug 27, 2026

Choose a reason for hiding this comment

Flue history never hydrates the panel

Uh oh!

cursor Bot Aug 27, 2026

Choose a reason for hiding this comment

Clear chat keeps the Flue conversation

Uh oh!

cursor Bot Aug 27, 2026

Choose a reason for hiding this comment

History failures look like empty chats

Uh oh!

codspeed-hq Bot commented Aug 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will not alter performance

Footnotes

Uh oh!

github-actions Bot commented Aug 27, 2026

Benchmark results

@rust/hash-graph-benches – Integrations

policy_resolution_large

policy_resolution_medium

policy_resolution_none

policy_resolution_small

read_scaling_complete

read_scaling_linkless

representative_read_entity

representative_read_entity_type

representative_read_multiple_entities

scenarios

Uh oh!

cursor Bot left a comment

Choose a reason for hiding this comment

Uh oh!

cursor Bot Aug 27, 2026

Choose a reason for hiding this comment

Pending client tools hydrate as complete

Uh oh!

cursor Bot Aug 27, 2026

Choose a reason for hiding this comment

Turns drop the idempotency key

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

1 participant

vercel Bot commented Aug 27, 2026 •

edited

Loading

lunelson commented Aug 27, 2026 •

edited

Loading

cursor Bot commented Aug 27, 2026 •

edited

Loading

codspeed-hq Bot commented Aug 27, 2026 •

edited

Loading