backpass 0.1.12 → 0.1.13

package/README.md

@@ -158,8 +158,9 @@ strict JSON: which instructions helped, which were violated, and what mistakes n
 instruction covers. Every negative carries a class - `harm` (following the instruction
 caused damage), `non-compliance` (the agent ignored it), or `irrelevant` - because those
 argue for opposite fates: harm argues against an instruction, non-compliance argues for
-reinforcing it. Every gap carries a domain - `project` for this repository's own
-engineering, `orchestration` for the task-management layer around the session - and the
+reinforcing it. Every gap carries a domain - `orchestration` when the mistake was caused
+not by this repository but by an external agent harness or tooling that orchestrated the
+task, `project` for every other mistake - and the
 analysis is shown the ledger's open gaps so it can cite an existing gap id instead of
 coining a paraphrase of it.
 
@@ -193,8 +194,9 @@ consolidation call sees the full open gap set and merges entries that describe t
 mistake. That second judgment is what lets two sightings of a brand-new gap in the same
 run's parallel fan-out corroborate. A failed consolidation call degrades the run to
 lexical identity and says so; it never aborts. Orchestration-domain gaps are counted and
-reported but never cluster: mistakes about the task harness around a session do not
-become instructions in the project's memory file.
+reported but never cluster: a mistake caused not by this repository but by the external
+agent harness or tooling that orchestrated a session does not become an instruction in the
+project's memory file.
 
 Only evidence judged against the _current_ memory-file set hash is folded into a proposal. A
 transcript that fell out of this run's sample - the time window, `maxTranscripts`, or the

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "backpass",
-  "version": "0.1.12",
+  "version": "0.1.13",
   "packageManager": "pnpm@11.5.0",
   "description": "Gradient descent for your agent memory - analyzes past agent session transcripts and proposes evidence-backed edits to AGENTS.md / CLAUDE.md",
   "type": "module",

package/src/fold.js

@@ -91,9 +91,9 @@ export function foldEvidence(evidenceRecords, { minGapEvidence = 2, memoryFile =
     }
   }
 
-  // Orchestration-domain sightings are about the task-management layer around the
-  // session, not this repository's engineering; they are counted for legibility but
-  // never cluster, so they can never corroborate into a project proposal.
+  // Orchestration-domain sightings are mistakes caused not by this repository but by the
+  // external agent harness or tooling that orchestrated the session; they are counted for
+  // legibility but never cluster, so they can never corroborate into a project proposal.
   const allObservations = gapObservations ?? recordObservations;
   const projectObservations = allObservations.filter((obs) => obs?.domain !== "orchestration");
   const orchestrationGapSightings = allObservations.length - projectObservations.length;
@@ -230,8 +230,8 @@ export function renderEvidenceForPrompt(summary) {
   lines.push("### Gap clusters (mistakes no current instruction covers)");
   if (summary.totals.orchestrationGapSightings) {
     lines.push(
-      `- ${summary.totals.orchestrationGapSightings} orchestration-domain sighting(s) about the ` +
-        `task-management layer were excluded; they never enter this repository's memory file`,
+      `- ${summary.totals.orchestrationGapSightings} orchestration-domain sighting(s) caused by the ` +
+        `orchestrating harness or tooling were excluded; they never enter this repository's memory file`,
     );
   }
   if (!summary.gaps.length) {

package/src/gap-ledger.js

@@ -26,9 +26,9 @@ import { sha256 } from "./state.js";
  *    merged by the pre-synthesis consolidation pass (`mergeGapEntries`, driven by
  *    `src/consolidate.js`), which is what lets two same-run parallel sightings - neither
  *    of which could cite the other - still corroborate.
- *  - Every observation carries the `domain` the analysis judged: `project` for mistakes
- *    about this repository's own engineering, `orchestration` for mistakes about the
- *    task-management layer around it (briefs, scout scope, status records, approvals).
+ *  - Every observation carries the `domain` the analysis judged: `orchestration` when the
+ *    mistake was not caused by this repository but by an external agent harness or tooling
+ *    that orchestrated the task, `project` for every other mistake.
  *    Orchestration sightings are recorded for legibility but never counted toward
  *    corroboration and never surface in a proposal; a missing domain counts as project,
  *    so evidence from before the field existed keeps its old behavior.

package/src/prompts/analysis.md

@@ -53,12 +53,12 @@ Rules, in order of importance:
    or violated the instruction - evidence the instruction failed to steer, which argues
    for reinforcing it, never for deleting it. `irrelevant`: on inspection the moment
    does not actually bear on this instruction. Never report a skipped rule as `harm`.
-4. **`domain` states whose mistake a gap is.** `project`: about this repository's own
-   engineering - its code, tests, build, docs, releases, conventions. `orchestration`:
-   about the task-management layer around the session - task briefs and their scope
-   (scout/read-only rules), status reporting to a supervisor, approval and authorization
-   flows, delivery-lifecycle process imposed from outside the repository. Orchestration
-   gaps are counted but never proposed into this repository's memory file.
+4. **`domain` states what caused a gap.** A gap is `orchestration` when the mistake was
+   not caused by this repository, but by an external agent harness or tooling that
+   orchestrated the task (a task brief, a supervisor's process, the harness itself - by
+   way of illustration only, not a list to match against); every other gap is `project`.
+   Ask the causal question, not which category the wording resembles. Orchestration gaps
+   are counted but never proposed into this repository's memory file.
 5. **Do not confabulate influence.** Only call something positive when the trace shows
    the agent doing the specific thing the instruction asks for. An outcome that would
    have happened anyway is not evidence.

package/src/prompts/synthesis.md

@@ -81,7 +81,7 @@ analyzed sessions in which an instruction drew any evidence at all.
 5. **Every edit must be backed by at least one verbatim quote** from the evidence. You
    will attach the quotes in the next step, so only make changes you can back.
 6. **Budget:** {{BUDGET_RULE}}
-7. Prefer extracting a long, narrow, crisply-triggered section over deleting anything:
+7. You can extract a long, narrow, crisply-triggered section instead of deleting it:
    extraction frees the same always-loaded tokens and loses nothing.
 8. Change only `./{{MEMORY_PATH}}` and files under `./{{SKILLS_DIR}}/`. Never delete a
    file. Do not create notes, scripts, or scratch files.
@@ -94,7 +94,7 @@ analyzed sessions in which an instruction drew any evidence at all.
 | Conditional / narrow     | **skill** (the description is the condition) | deletion candidate |
 
 A skill's description is always loaded and its body is free until triggered, so moving a
-long, narrow, crisply-triggered section into a skill is nearly pure budget profit.
+section into a skill trades its always-loaded cost for that one description line.
 
 **Skill descriptions are weights too.** If the evidence shows an agent lacked knowledge
 an existing skill already contains, that is a failed trigger: rewrite that skill's