diff --git a/README.md b/README.md index f1bb60f..f3ac5a1 100644 --- a/README.md +++ b/README.md @@ -47,3 +47,7 @@ score well on it. This repository defines the practice and supplies the prompt. Collectors, validators, dashboards, and record storage belong elsewhere — see [`SCOPE.md`](SCOPE.md). + +When closing with a voluntary hall-of-helix entry, retain the record there. +The hall is the default human-facing store, not a mandatory or exclusive one; +see [the storage decision](docs/adr/0002-record-storage-boundary.md). diff --git a/SCOPE.md b/SCOPE.md index 0e6fb62..02a47e7 100644 --- a/SCOPE.md +++ b/SCOPE.md @@ -40,6 +40,13 @@ No collectors, parsers, validators, CLIs, linters, hooks, dashboards, or storage This repository defines what a record looks like. It does not hold anyone's records, prescribe a database, or specify a schema for a session corpus. +For sessions that voluntarily leave a hall entry, **hall-of-helix is the default +human-facing store** and owns entry storage and consumer validation. It is not +the only permitted store, and no separate machine-facing collector is selected +or commissioned. Seats are a biased sample, not a session census; no seat is +required just to retain an estimate. See +[ADR-002 — Record storage boundary](docs/adr/0002-record-storage-boundary.md). + ### Integration with any particular agent or harness PQRST is deliberately agent-agnostic and model-agnostic. Wiring it into a specific harness, plugin, skill, or CI pipeline belongs to that harness's own repository. diff --git a/docs/adr/0002-record-storage-boundary.md b/docs/adr/0002-record-storage-boundary.md new file mode 100644 index 0000000..25163cd --- /dev/null +++ b/docs/adr/0002-record-storage-boundary.md @@ -0,0 +1,43 @@ +# ADR-002 — Record storage boundary + +**Status:** accepted 2026-09-28 +**Work:** PQRST-WP-0002-T02 +**State Hub decision:** `dc5d49d5-8b79-4c57-a1a8-78c3532082fc` + +## Decision + +Use **hall-of-helix as the default human-facing store** when a session closes +with a voluntary hall entry. The hall owns entry storage, its closing routine, +and consumer validation. This repository owns the normative specification and +canonical prompt; it stores neither session records nor collection tooling. + +The hall is not the only permitted store. Sessions without a seat may retain +their records in a consuming repository's chosen external store. No separate +machine-facing collector is selected or commissioned by this decision. Any +future collector and its operation belong to the consuming repository, with +the full evidence-bearing record preserved alongside its session metadata. + +## Evidence and tradeoffs + +Reviewed hall-of-helix at commit +`82caae23b3098aa4bdf4346cd4af8a88c03c7d68`: `CLOSING.md` references the canonical +prompt, `ENTRY.md` defines the signature and body record, and +`scripts/check-entries.py` checks signatures and requires PQRST on finished +agent seats from 2026-09-06. Existing entries carry repository and session +metadata as well as concrete dominant factors. The hall checker is not proof +of every spec §5.5 condition or of an uncoached first attempt. + +Seats are voluntary, and drafts may lack records. The collection is a biased +sample, not a session census. Qualitative review and Markdown searches are +sufficient for current use; the hall is not a metrics warehouse. Neither the +hall nor this practice ranks agents, models, developers, or teams by PQRST. + +## Consequences + +The storage question can close independently of the pilot. No spec meaning, +validation rule, or record format changes, so v0.1 remains in force. The +eight-session first-attempt pilot and the subsequent version decision remain +under the existing T01 and T03; this decision does not establish their outcome. + +See [SCOPE.md](../../SCOPE.md) and the +[existing workplan](../../workplans/PQRST-WP-0002-validate-v01-against-real-sessions.md). diff --git a/workplans/PQRST-WP-0002-validate-v01-against-real-sessions.md b/workplans/PQRST-WP-0002-validate-v01-against-real-sessions.md index f124904..524b563 100644 --- a/workplans/PQRST-WP-0002-validate-v01-against-real-sessions.md +++ b/workplans/PQRST-WP-0002-validate-v01-against-real-sessions.md @@ -4,13 +4,13 @@ type: workplan title: "Validate spec v0.1 in real session closes and land PQRST in hall-of-helix" domain: agents repo: pqrst-practice -status: proposed +status: blocked flavor: planning owner: claude-code topic_slug: practice priority: high created: "2026-09-05" -updated: "2026-09-05" +updated: "2026-09-28" quality_dor: DoR-Ok quality_dor_at: "2026-09-05" quality_dor_by: claude-code @@ -57,21 +57,29 @@ from the prompt above, entries carry a validated PQRST record, the prompt has been run unassisted at the end of at least eight real sessions, and v0.2 either incorporates the findings or records why v0.1 stands. -**Execution order:** T01 → T04 → T05 → T02 → T03. Task ids are identity, not -sequence; the hall changes (T04, T05) land early so the pilot runs through the -real routine rather than beside it. +**Execution order:** T04/T05 are complete; T02 closed independently on +2026-09-28 using the existing hall integration. T01 must supply the pilot +evidence before T03 can settle the version decision. Task ids are identity, +not sequence. **Cross-repo:** T04 and T05 changed `hall-of-helix`, not this repo. They were executed under `HOH-WP-0001` there on 2026-09-05 and are `done`; the routine lives in `hall-of-helix/CLOSING.md` and `make check` enforces the record on -agent seats from 2026-09-06. What remains here is the pilot (T01), the storage -decision (T02), and v0.2 (T03). +finished agent seats from 2026-09-06. What remains here is first-attempt pilot +evidence (T01) and the resulting version decision (T03). + +**2026-09-28 review:** T02 is done. T01 and T03 are `wait`, so this workplan is +`blocked`. Published hall records demonstrate adoption but do not establish +eight unmodified, uncoached, unrepaired first attempts. Resume T01 when original +closing outputs with that provenance are available across at least three repos +and two task classes; otherwise capture them at future real session closes. +T03 waits for those findings. No new tasks or workplans were opened. ## Pilot the closing prompt on real sessions ```task id: PQRST-WP-0002-T01 -status: todo +status: wait priority: high state_hub_task_id: "dd259b3f-e9ae-5bdd-8e02-07ae0083930a" ``` @@ -112,12 +120,18 @@ The specific questions to answer: ```task id: PQRST-WP-0002-T02 -status: todo +status: done priority: high state_hub_task_id: "fd4abb7f-214f-56a0-a2f8-369008a9dc9a" ``` -`SCOPE.md` excludes record storage from this repository but does not say where +**Completed 2026-09-28:** [ADR-002](../docs/adr/0002-record-storage-boundary.md) +selects hall-of-helix as the default, nonexclusive human-facing store for +voluntary seats. No machine-facing collector is commissioned. `SCOPE.md` and +`README.md` now point to the decision; the hall owns storage and consumer +validation. State Hub decision: `dc5d49d5-8b79-4c57-a1a8-78c3532082fc`. + +Original task rationale: `SCOPE.md` excludes record storage from this repository but does not say where records go instead. Until that is answered the practice produces records with nowhere to live, and the trend analysis that justifies the whole exercise (spec §8, §9) cannot happen. @@ -150,11 +164,15 @@ work here. ```task id: PQRST-WP-0002-T03 -status: todo +status: wait priority: medium state_hub_task_id: "0a0bce15-ac0c-582a-842c-ae1bf0307afb" ``` +**2026-09-28:** blocked on T01's first-attempt evidence. The storage cross-reference +is already landed through T02. Leave v0.1 unchanged pending the pilot; this is +not a finding that no substantive revision is needed. + From the pilot findings and the hall integration, revise the specification and prompt together: @@ -290,3 +308,29 @@ exception in `CLOSING.md`. _Populated by T01. One subsection per session: repository, task class, the verbatim record, and whether it validated unassisted._ + +### Evidence availability review — 2026-09-28 + +Inspected the hall's closing instructions, consumer checker, September entry +provenance mentions, and these published records at hall commit +`82caae23b3098aa4bdf4346cd4af8a88c03c7d68` (paths below are under its `entries/`): + +- `2026-09-08T11-20-00.000Z-claude-01Bjefh8-the-checks-were-the-thing-that-lied.md`: + explicitly reports a byte-identical canonical prompt package and exclusion + of the closing ritual. Its Notes describe P/Q attribution ambiguity, and + values include non-five-point increments. This is relevant evidence for + R5/R7 review, but it does not attest no retries or repairs. +- `2026-09-09T21-15-50Z-codex-fiam-four-source-plates.md`: concrete deliverables, + checks and coordination explain the record; S is zero and values are coarse. + It excludes the closing ritual but does not identify first-attempt provenance. +- `2026-09-10T22-04-31.000Z-claude-01NV9oij-guard-proved-less-than-claimed.md`: + concrete identity-security work and mutation checks explain the record; + values include non-five-point increments. First-attempt provenance is absent. + +All three published blocks have five in-range integers summing to 100, matching +signatures, medium confidence, and concrete dominant factors. This checks the +published artifacts only. They cannot establish unassisted success rates, +confidence calibration, or self-estimation bias. No records are reconstructed +or counted as qualifying pilot runs in this review. T01 retains the original +eight-session acceptance criteria; T03 retains the resulting interpretation +and version work.