FOUND BY E2E, NOT BY REASONING. The previous commit's design note asserted the provenance
receipt "never reaches the user: it is appended to the history copy, not the reply." That
was FALSE, and only running the thing showed it. On the first live run against the built
DMG brain, two agentic turns out of two came back with
"Your favourite colour is chartreuse and your project is called Perihelion.
[[RECEIPT - recorded by the soul, not written by the model: no tools ran on this turn.]]"
— the receipt in the user-visible reply.
MECHANISM: the receipt is stored inside the assistant turn, and the agentic path replays
history VERBATIM as Anthropic message objects. So the model sees its own previous answers
ending in [[RECEIPT ...]] and does the obvious thing — it imitates the format and signs the
next answer the same way. The plain path did NOT leak, which is the tell: there, history is
rendered into the SYSTEM prompt as labelled lines rather than replayed as assistant turns,
and a model imitates its own turns far more readily than a transcript.
FIX, two layers, because one of them is not a guarantee:
- receipt_rule() names the marker in both system prompts (plain and agentic): these lines
are written by the system, read them as evidence, never write one. Reduces occurrence.
- receipt_strip() truncates any [[RECEIPT ...]] out of model output before it becomes the
reply — plain path in layered_generate, agentic path on final_text in agentic_loop.
Deterministic. A guard that depends on the model choosing to obey is exactly the class of
thing round 8 exists to stop shipping, so the instruction is the optimisation and the
strip is the guarantee.
Placed ABOVE agentic_loop's empty-check on purpose: a turn whose entire output was an
imitated receipt has produced no answer, and must be reported as no answer.
The receipt stays in HISTORY, which is the whole point and is proven to work: asked "What
source did you use for that?" one turn after a live web_search, this brain answered
"I used Weather Underground (https://www.wunderground.com/weather/is/reykjav%C3%ADk) for the
current temperature in Reykjavik" — a real source, no apology. That is the false confession
dead, and it is dead BECAUSE the model can read the receipt.
BUILT: 887,112 bytes, sha256 54a2eff84d4fa44f8d2db6781dcf225df4b5075a8ec5f1058018bd40cc1af10b
BUG-PLAINCHAT-1 miscompile guard: zero sites.
Refs neuron#109
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
DESIGN FIT: three of round 7's five defects share ONE root — the conversation-history
layer persists only {role, content}, discarding tool provenance, session scoping, and the
distinction between a real user turn and an internal utility call. Fixes A and B RESTORE
Will's design rather than extend it: his agentic path already scopes history per session,
the plain path never got it, and his own source carries the TODO admitting the resulting
race (chat.el, handle_chat: "process-global key; concurrent /api/chat requests without
session_id race on this read-append-write"). Fix C repairs one join Will wrote that was
correct for a year and one we added last week. E1/E2 are ours.
FIX A — tool provenance in history (kills the FALSE CONFESSION)
Root cause, EXECUTED-verified: handle_chat_agentic recorded turns via hist_append, which
emits {"role","content"} only. server_tool_use blocks, web_search_tool_result blocks and
every citation were discarded, then replayed as text. On the next turn the model saw a
data-rich answer with zero evidence a search had happened, and its own permanent rule
("never describe a search you did not perform") left one conclusion available: that it
had fabricated the data. It apologised for a search it HAD run — four independent lines
of evidence confirm the search was real. The defect is not the model's honesty. It is
that we deleted the evidence and then asked it to account for itself.
Change: agentic_loop accumulates the source URLs it already walks past (citations and
web_search_tool_result content) and returns them as "sources"; handle_chat_agentic folds
tools_used + sources into a receipt line stored WITH the assistant turn. Receipts are
unconditional — a negative receipt ("no tools ran") is the other half of the guarantee,
because "no evidence of a tool" and "evidence of no tool" were previously identical in
the transcript. conv_history_block splits the receipt off before snipping so a long
answer cannot truncate away the evidence. The user never sees it: it is appended to the
history copy, not the reply.
FIX B — one history key for both paths (kills the BLANK STARE)
Root cause, EXECUTED-verified: the agentic path keyed history on session_hist_<id>; the
plain path was hard-wired to the process-global conv_history and never read session_id.
One conversation, two buckets. Proven in the guest engram: the scoped node held exactly
two turns starting at "Try again" while the earlier exchanges sat unscoped.
Change: conv_hist_key/conv_hist_label are now the single definition, used by BOTH paths;
session_id is threaded route -> layered_cycle -> layered_generate / conv_history_record.
The 2-line fallback (plain path reads the agentic key) was REJECTED: it keeps the
process-global bucket as a live write target, which is the bleed the TODO describes.
Also found and closed while threading: layered_cycle read session_id from the state key
"current_session_id", which is read here and WRITTEN NOWHERE in the entire source. It
was unconditionally "", so TODO(reliability #4) — per-session steward continuity — was
dead code that could never fire. It fires now.
LAZY SESSION, decided explicitly: we create the session EAGERLY at the door (app half,
ui#223) rather than migrating orphaned turns. Migration would copy the CONTENTS of a
process-global bucket, possibly another conversation's, into a named session — the bleed,
performed deliberately. Eager creation makes the situation impossible instead. Migration
is deliberately not implemented and must not be added without solving provenance first.
FIX C — the two text-join seams ("to.Good", byte-verified 0x77 0x2e 0x47)
Two bare `+` joins, written a year apart, had drifted into two answers to one question:
within-response block joins (Will's, 2026-05-03, latent until server-side web_search
began interleaving non-text blocks) and across-round joins (ours, 62af564).
Change: one named rule, text_join_sep, at both sites. NOT a blanket separator — a cited
answer splits MID-SENTENCE ("The current temperature is " + "86°F" + ", with "), so a
blanket separator shatters every sourced sentence. The rule takes the one bit that
distinguishes the cases: whether a NON-TEXT block intervened. Hoisting it also makes the
fix verifiable in the shipped binary, which an inline `+` is not.
FIX E1 — utility generations stay out of the transcript
Title generation ("Write a 3-6 word title...") and insight passes ran down the same plain
door as a real message and were recorded as if the user had typed them; the same calls are
the "model":"unknown" rows in usage.jsonl. is_utility_request reads an explicit utility
flag from the app, with the __title__/__insight__ id prefixes as a fallback for older
clients. Answered normally, never recorded.
FIX E2 — OPERATOR IDENTITY is scoped to tool-capable turns
The block (env USER/HOME, closing "This is a hard rule") was prepended to EVERY system
prompt including chat mode. On a Tools:Off turn there is no filesystem in reach, so it
governed nothing and merely supplied the loudest fact in the prompt — which is why the
model opened a fresh conversation with "You're test, on your machine at /Users/test".
Hoisted to operator_identity_block() and gated on !chat_mode. Unchanged wherever a file
or command tool can actually be reached.
ALSO: agentic_loop's per-session history persist had a second hand-rolled copy of
conv_history_persist with a different label expression, different salience scores and
different tags for the same node. Since both now derive the label from conv_hist_label and
engram_node_full upserts by label, two score policies were writing one node. Collapsed to
one writer.
BUILD NOTE: dist/elp-c-decls.h is force-included by the documented link recipe and carried
the OLD C arities, so it is updated here. This is the build-support header, NOT the stale
generated dist/soul.c — no dist/*.c was read or edited; all engine changes are .el source.
chat.elh/soul.elh are committed because a first-pass build against the old signatures FAILS
(measured); the other regenerated headers are reverted as unrelated churn.
BUILT: 887,000 bytes, sha256 d632b061ad75269d6adeb52578d030eaf49e895d91289d7f946b19c08450d728
Zero el_str_concat(<int>, str_len(...)) sites (the BUG-PLAINCHAT-1 miscompile guard).
web_search_20250305 and disable_parallel_tool_use both still present — PR #108's web search
and the ADR-0005 stopgap are intact.
Refs neuron#109 (builds on it), neuron#78 (Receipt Contract — the real fix A is a stopgap for)
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
awareness_run's while-loop ran outside any request arena, so every
allocation in every 1s tick (search JSON, heartbeat payloads, curiosity
activations) was treated as permanent by the runtime — 7.5GB RSS in
under a minute. Bracket each iteration with el_arena_push/el_arena_pop
(same pattern the compiler emits for scoped blocks; state_set/state_get
persist separately via el_strdup_persist and are unaffected).
dist/soul.c carries the same change hand-patched at the compiled
awareness_run site — elc is currently unsafe to run locally (pathological
memory on sessions.el), so the generated C was patched to match the
source, verified line-for-line against the compiler's own conventions.
MUST be paired with el repo PR #64 (el_strdup_persist for stored engram
fields): per-tick arena reclamation widens the write-corruption window
without it. Verified together: 5h live soak on the recovered production
snapshot, flat RSS, write-field-integrity clean.
Note: dist/soul.c still needs a full elc regen to pick up PR #73's
source changes (consent tiers) — tracked separately; this patch does not
regress that (those changes were never in dist).
- PR #60: inject operator home dir into system prompt (#30)
Adds OPERATOR IDENTITY section so the LLM correctly resolves
'my files/notes/desktop' to the actual running user's $HOME.
Prevents identity confusion between imprint author and operator.
- PR #61: plan-mode endpoint POST /api/chat {mode:'plan'} (#27)
Adds handle_chat_plan — returns {steps:[{id,title,detail}]} JSON.
Wired into all three /api/chat route handlers. Grounds the plan
via engram_compile (same as agentic path) for context awareness.
dist changes:
- soul.c: both PRs compiled in; build_system_prompt updated to
2-param signature (ctx, chat_mode); handle_chat_plan added
- chat.c/routes.c/chat.elh: individual module outputs updated
- elp-c-decls.h: remove stale 1-param build_system_prompt decl,
add handle_chat_plan declaration
- soul.elh.c: new soul header declarations file (from PR #60)
Compile verified: cc -O2 -DHAVE_CURL soul.c el_runtime.c -lcurl
Binary: 805K arm64, smoke test passes (port in use = expected).
- sessions.el: new sessions module with session management and approval gate
- routes.el: wire /api/sessions routes (list, get, create, approve, tool_result)
- chat.el: thread-aware activation — short messages anchor to last reply
before engram compilation so follow-ups stay on-topic
- chat.el: agentic path tracks per-session history (session_hist_{id})
instead of shared conv_history, seeding each turn with prior context
- chat.el: add call_neuron_mcp, dispatch_tool, is_builtin_tool, next_bridge_id
agentic_loop, bridge_save, agentic_resume, handle_tool_result
- dist/soul: rebuild with all of the above
neuron-api.el is a new first-class El module that implements all Neuron
cognitive API handlers natively — no HTTP round-trips, no MCP wrapper,
direct engram builtin calls. All capabilities that previously lived in
the MCP wrapper adapter now live here in the soul.
Handlers: begin_session, compile_ctx, remember, recall, search_knowledge,
browse_knowledge, capture_knowledge, evolve_knowledge, promote_knowledge,
browse_processes, define_process, log_state_event, list_state_events,
inspect_config, tune_config, inspect_graph, link_entities, list_typed,
consolidate.
Routes wired in routes.el under /api/neuron/* (GET + POST).
Also compiles all loop-1/loop-2 .el source changes into dist/*.c and
rebuilds the binary. memory.elh and neuron-api.elh updated with new exports.
Adds handle_dharma_room_turn_agentic to chat.el — same full tool loop
as handle_chat_agentic but reads transcript directly (not message), and
returns {response, cgi_id, tools_used} to match the dharma room shape.
Registers dharma_room_turn_agentic as a new event type in routes.el so
the studio can dispatch @neuron turns through this dedicated path.
- soul.el: SOUL_CGI_ID, SOUL_ENGRAM_PATH, SOUL_IDENTITY env vars;
state_set("soul_snapshot_path") so callers can find it; only call
init_soul_edges() when cgi_id == "ntn-genesis"
- chat.el: handle_dharma_room_turn — soul builds its own context from its
own engram, assembles system prompt, calls LLM, persists episodic memory;
also fix is_new_tool scoping bug in handle_chat_agentic (use has_tool)
- routes.el: wire dharma_room_turn event type before chat_as_soul branch
- rebuild dist/neuron: handle_dharma_room_turn now compiled in
Replace scan-by-offset fallback with engram_get_node_json calls for the
known high-salience identity nodes (family, origin). Offset-based scanning
is order-dependent and unreliable; direct ID fetch is stable regardless of
snapshot position. Ensures biographical context (Fox, Bobby, etc.) is
always in the system prompt when vector search returns nothing.
Routes a new event_type "chat_as_soul" through dharma/recv. The Studio
preassembles the system_prompt + transcript and dispatches per-speaker;
the soul-binary just performs the LLM call as the requested speaker_slug.
No engram_compile here — each soul has its own engram (88xx) and the
Studio queries it before composing the prompt.
Also: track the previously-untracked split source modules (chat, routes,
memory, awareness, studio) and add build.sh so the binary can be rebuilt
without the studio’s concat trick. elb resolves the import graph and
emits one .c per .el; we link them together with cc. dist/soul-el now
points at dist/neuron via symlink (matching the launchctl plist).
Tim Lingo
will.anderson