kyle/boxmaker
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
425 Commits 3 Branches 0 Tags
fb2afa19798047b0668bbc2c5adc5c6acda9955a
Commit Graph
197 Commits
Author SHA1 Message Date
kyleandornith-1.5-35b-a3b fb2afa1979 bxctl memory sync: the Honcho ledger's counts and failed items 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 13:22:26 -07:00
kyleandornith-1.5-35b-a3b fb05b26462 memoryd: the Honcho worker sends conclusions, their deletions and fleet notes 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 13:11:51 -07:00
kyleandornith-1.5-35b-a3b 4ebfc36c1a memoryd: memory writes decided into the Honcho ledger as they are journalled 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 12:46:18 -07:00
kyleandornith-1.5-35b-a3b b37eaef390 memoryd: the Honcho worker sends decided turns, waits out outages, retries undecided turns 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 12:30:49 -07:00
kyleandornith-1.5-35b-a3b aabfbc2f0a memoryd: a Honcho client per workspace, and its workspace, peers and sessions made sure of 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 12:13:48 -07:00
kyleandornith-1.5-35b-a3b 28df2a0e3b memoryd: a reported turn decided after loopd has its answer, from brokerd and the log 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 11:59:34 -07:00
kyleandornith-1.5-35b-a3b a339eb58b6 memoryd: where a finished turn goes in Honcho, decided from brokerd and the log 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 11:41:38 -07:00
kyleandornith-1.5-35b-a3b 8aaeb74f36 memoryd: the Honcho ledger, decided, sent and failed, read through at start 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 11:33:49 -07:00
kyleandornith-1.5-35b-a3b 129a9eadc3 memoryd: [honcho] and its keys checked at start, and brokerd's state socket 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 11:20:10 -07:00
kyleandornith-1.5-35b-a3b 9b4e9807b2 brokerd: state.sock answers memoryd's question about a session's state 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 11:05:42 -07:00
kyleandornith-1.5-35b-a3b 5199bd7149 bxctl: chat --speaker NAME, the speaker on every turn of the chat 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 10:57:11 -07:00
kyleandornith-1.5-35b-a3b fae12c3fe2 gatewayd: a turn's speaker, the username of the one person who wrote its burst 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 10:51:17 -07:00
kyleandornith-1.5-35b-a3b 4866353ac9 loopd: a turn's speaker in the log, its number in SessionUpdated 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 10:33:36 -07:00
kyleandornith-1.5-35b-a3b c798714af1 bxcore: a turn's speaker, SessionUpdated's turn, the session-state messages 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-28 10:27:28 -07:00
kyleandClaude Opus 5.5 027146f359 M5c-1 plan: README, tasks 07 to 15, the owner's decisions, the known flaky test; M5b finding 8 closed 
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-28 10:00:59 -07:00
kyleandClaude Opus 5.5 d423d87115 docs: the M5b device check, passed on straylight 
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-27 19:08:42 -07:00
kyleandClaude Opus 5.5 6f243ad779 docs: M5b review finding 10, the tag check against inferproxy; tip T49 
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-27 19:08:42 -07:00
kyleandClaude Opus 5.5 1fb0ec8a06 docs: the M5b review, and tips T45 to T48 
The review of tasks 01 to 14 in the implementer log (task 14's row
renamed to its file's name), and four tips for writing tasks: what to
do with a skeleton's comments, golden files for scanners, naming the
log row, and running the implementer without plugins.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 27a6354a80 docs: runbook: sanitizer-rules, how to change a sanitizer rule 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 7cbc1d7342 brokerd, memoryd, gatewayd: a lock file keeps each to one instance 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b fd0437a032 bxcore, loopd, inferproxy: a lock file keeps each to one instance 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 54f102bbe1 bxctl: sanitize check shows which sanitizer rules a text breaks 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 648469cd67 gatewayd: images, and links in untrusted replies, are posted as code 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 0d3048c3c4 bxcore, loopd: TurnDone says whether the session has read untrusted content 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 91992b9743 loopd: the self-test checks the model's tags against the tokenizer 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b b4d631e249 loopd: neutralise user messages, tool results and the memory block 
Text from outside the model (the user's message, every tool result, and the
memory block) is neutralised the moment it is first logged, so every later
request replays the same bytes and a tag in that text cannot reach the model as
one of its own control tokens. The model's own output, replay, and the system
prompt are left byte-for-byte.

Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b ca8b50cd69 memoryd: refuse and flag memory writes through bxcore::sanitize 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b b3e9af7586 bxcore: sanitize::flag, text shaped like an attack 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 41d425f73e bxcore: sanitize::refuse, what a memory write may never hold 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b 1817e7f117 bxcore: the model's tags and sanitize::neutralise 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b f594e0d9d4 bxcore, gatewayd, loopd: the secret store moves to bxcore::secrets 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandornith-1.5-35b-a3b de60c9e13a bxcore, brokerd, loopd: an invalid-arguments denial says why 
Co-Authored-By: ornith-1.5-35b-a3b <agent@wntrmute.dev>
 2026-09-27 19:08:42 -07:00
kyleandClaude Opus 5.5 e41a204902 M5a review: tasks 01 to 22 and the device check; tips I16, I17, T34 to T40 
Two review entries in the implementer log: tasks 01 to 15 (the header
forgery, review ids, the runbook entries, log rows written from memory)
and tasks 16 to 22 with the device check on straylight (the ledger
raising state by the grant's label, the note answers, the stale target/
that stalled task 21, the clippy 1.95 stop, the fake podman race). Two
short rules go into AGENTS.md: stop after one check when a test outside
the task fails, and write the log row from the diff.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-25 21:04:23 -07:00
kyle e963ef72bb memoryd: note answers name notes.md and the file that was not changed 2026-09-25 20:32:29 -07:00
kyleandClaude Opus 5.5 7b95937c88 brokerd: the ledger raises a session's state by the stricter label 
M5a task 21. The work is Ornith's, from the stopped run; the design
model checked it against the task file and committed it after finding
that the gate's failure came from its own reference worktree's stale
test binaries in the shared target/ (fixed with cargo clean; 944 tests).

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-25 20:20:48 -07:00
kyle 34dbbb27be memoryd: a staged write's id counts every write staged that day 2026-09-25 16:41:44 -07:00
kyle 81107bc5c6 proto, memoryd: no line of an entry may start with §, and headers are exact 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle 3640d32de0 memoryd, proto, brokerd, bxctl: repeated error lines written once 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle b98f373ead memoryd: the day and the month of a time with get, never a slice 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyleandClaude Opus 5.5 7d5c5a8d85 gate: pass on macOS: tests use /tmp, gatewayd finds its uid without /proc 
M5a task 16. The work is Ornith's, from the stopped run; the design
model checked it against the task file and committed it once the
memoryd lint that stopped the gate was fixed (44bdcbd).

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-25 16:41:44 -07:00
kyle e763ad715e log: M5a/16-macos-gate stopped on a pre-existing memoryd clippy failure 
The three task changes (.cargo/config.toml, copied toolkit shell test, and the three read_checked edits in gatewayd/secrets.rs) are done and all gatewayd/toolkit/brokerd tests pass. make gate does not reach them: it fails to compile memoryd on clippy collapsible_match at crates/memoryd/src/index.rs:224, which fails identically on the original committed code and is outside this task's scope. Per AGENTS.md point 4, committed only this log row.
 2026-09-25 16:41:44 -07:00
kyleandClaude Opus 5.5 fb65206a6c Implementer log: the Model column filled in as Ornith-1.5-35B-A3B 
Every row from M2a to M5a that read ? or Ornith, from the reviews' own
records of who ran each milestone. The DEBUG-HANDOFF row stays ? : no record
says which model it was.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-25 16:41:44 -07:00
kyleandClaude Opus 5.5 6f093fcd8e M5a task 15: the starter soul and device checks with memoryd, finished 
The session did the task and stopped on make verify-device. The cause
was not the server: run-plan.sh and the device tests both read
BOXMAKER_MODEL, and the run's value (OpenCode's provider-prefixed id)
reached llama-server as the model name. With it unset, verify-device
passes 6 of 6. The copied files are identical to the plan's, and the
gate prints gate: ok, 936 tests.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-25 16:41:44 -07:00
kyle 9f57cfc381 M5a: the starter soul.md, and the device checks with memoryd running 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle 0a1c206d50 bxctl: reindex, and memory log, undo, pending, accept, reject 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle 15ee4b8e3d loopd: the memory block from memoryd, recall and memory as core tools 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle e167fc871b brokerd: memory calls to memoryd with the session's state; labels only stricter 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle fcfd590687 brokerd: recall and memory, decided under built-in grants 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle 0c8d136c6b memoryd: serve, three sockets and one lock over the files and the index 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00
kyle 49af057c81 memoryd: recall, with source tags and the recall log 
Implemented-By: OpenCode session (model recorded in docs/implementer-log.md)
 2026-09-25 16:41:44 -07:00