coulomb-research/workplans/CR-WP-0004-claim-evidence-discipline.md

---
id: CR-WP-0004
type: workplan
title: "M2: introduce minimal claim/evidence discipline for CR-000"
domain: communication
repo: coulomb-research
status: finished
owner: codex
topic_slug: coulomb-social
created: "2026-09-29"
updated: "2026-09-29"
quality_dod: DoD-Ok
quality_dod_at: "2026-09-29"
quality_dod_by: codex
quality_dod_note: "Obtained genuine independent criticism (fresh, context-isolated agent instance) that found a real internal-consistency violation self-criticism missed; linked first evidence entries; built minimal read-only query prototype. Remaining residuals (organizationally-independent review, claim sharpening, provenance field) handed off to CR-WP-0005 rather than folded in."
origin: residual
origin_ref: CR-WP-0003
state_hub_workstream_id: "d4e08468-ef7f-5dab-af8a-aecd4e6dc415"
---

# M2: introduce minimal claim/evidence discipline for CR-000

Implement M2 of `seed/CoulombResearchMilestonePlan.md`: test whether the
Coulomb-native claim/evidence/criticism representation is actually useful
for conducting research (Q2 in
`research/programs/RP-000-researching-research/questions.md`), by linking
CR-000's existing claims to real evidence and criticism rather than leaving
`evidence.md` empty.

This workplan carries forward two residuals named explicitly in
`CR-WP-0003`'s self-criticism (`CR-000-X001`, `CR-000-X002` in
`criticism.md`):

1. **Independent criticism.** All criticism CR-000 has received so far
   (`CR-000-X001`, `CR-000-X002`) is self-authored by the same human/agent
   pair that wrote the claims under criticism. `CR-000-C004` (criticism can
   operate continuously) is not meaningfully tested until CR-000 receives
   criticism from a different actor.
2. **Document-centrism.** `CR-000-C002`/`CR-000-C003` (currently `contested`
   in `claims.md`) argue papers should not remain the canonical
   representation and that a graph representation can outperform
   document-centric publication, but CR-000's own artifacts are flat
   Markdown with no queryable graph behind the ID cross-references.

Inputs: `research/programs/RP-000-researching-research/papers/paper-000/`
(all files, especially `claims.md` and `criticism.md`),
`seed/CoulombResearchMilestonePlan.md` (M2 section), and
`docs/ResearchArtifactConvention.md`.

## Solicit and Record Independent Criticism

```task
id: CR-WP-0004-T01
status: done
priority: high
state_hub_task_id: "540baf89-2858-5edc-9c43-a78e6283ee88"
```

Obtain at least one criticism entry in `criticism.md` from an actor other
than the human/agent pair that authored `CR-WP-0002`/`CR-WP-0003` (e.g. a
human collaborator's review, or a differently-configured agent/model asked
to critique CR-000 independently, with its provenance recorded honestly per
`docs/ResearchArtifactConvention.md` §4). Do not simulate or fabricate an
"independent" critic using the same authoring context — that would not
resolve `CR-000-X002`, only relabel it. Update the resolution state of
`CR-000-X002` in `criticism.md` once independent criticism exists.

Done when at least one criticism entry's critic is verifiably distinct from
CR-000's own authors, and `claims.md`/`criticism.md` reflect its impact on
any claim it bears on.

## Collect and Link Real Evidence

```task
id: CR-WP-0004-T02
status: done
priority: high
state_hub_task_id: "368573d3-e917-5931-a42d-704b297e027e"
```

Populate `evidence.md` with real evidence entries (`CR-000-E001`, ...) that
bear on specific claims in `claims.md` — not placeholder or anticipated
evidence. Candidate evidence sources include: this program's own revision
history and workplan record as evidence about `CR-000-C001`/`CR-000-C005`
(human-AI participation, recursive self-improvement), and the outcome of
T01 as evidence about `CR-000-C004` (continuous criticism). Update each
linked claim's state (conjectural/supported/contested/etc.) in `claims.md`
to reflect what the evidence actually shows, including leaving a claim
`conjectural` if the evidence is inconclusive — do not inflate a claim's
state to `supported` on weak or single-instance evidence.

Done when at least the claims discussed in T01's outcome and in this task's
description have at least one linked evidence entry each, with claim states
updated accordingly.

## Prototype a Queryable Representation (or Explicitly Decline To)

```task
id: CR-WP-0004-T03
status: done
priority: medium
state_hub_task_id: "4bdc3ac4-5f91-55b3-93c3-726772520917"
```

Directly address `CR-000-X001`: either (a) build the smallest possible
prototype that makes claim/evidence/criticism relations actually
queryable (e.g. a script that loads all `CR-000-*` IDs and their stated
relations from the paper's Markdown files and answers a query like "what
challenges CR-000-C002" or "what evidence bears on CR-000-C003" without
manual grepping), consistent with the institutional-minimality principle
of building only what a concrete artifact demonstrates a need for; or (b)
if the facility's scale (one paper) does not yet justify even a minimal
prototype, record that reasoning explicitly as a revision-log entry and
leave `CR-000-X001` open with that rationale, rather than silently letting
it drop. Do not build a database or service — that remains out of scope
per the M0/M2 non-goals.

Done when `CR-000-X001` has either a working minimal prototype referenced
from `manuscript.md` §11, or an explicit, reasoned decision recorded in
`revision-log.md` for why not yet, with `criticism.md`'s resolution state
updated either way.

Closed: built `scripts/query_paper_graph.py` (read-only, stdlib-only,
parses relation fields across `claims.md`/`evidence.md`/`criticism.md` and
answers `--challenges`/`--bears-on`/`--id`/`--list` queries). Referenced
from `manuscript.md` §11 and §12. `CR-000-X001` left open — the prototype
is a query layer over flat files, not a real graph, and does not claim to
resolve the underlying criticism.

## Assess Representation Usefulness and Close Out M2

```task
id: CR-WP-0004-T04
status: done
priority: high
state_hub_task_id: "411e3a05-6f0d-5e66-9732-be453e89e4f9"
```

Using the outcomes of T01–T03, answer Q2
(`research/programs/RP-000-researching-research/questions.md`): was the
Coulomb-native claim/evidence/criticism representation useful for
conducting this research, or did it mainly add bookkeeping overhead? Record
the answer as a new section or revision to `manuscript.md`, update Q2's
status in `questions.md`, and update `claims.md` states for any claim T01–
T03 newly bear on. Run `python3 scripts/validate_research_registry.py` and
the canonical Repo Manager sync. Before closure, evaluate whether M3 (CR-001
registration) or a different next step is the appropriate live follow-up,
and hand it off per the residual convention.

Done when the M2 exit criterion (major manuscript claims are explicitly
addressable and linked to supporting or challenging evidence) is met for at
least the claims this workplan touched, and the registry validator passes.

Closed: Q2 answered as "mixed, leaning useful for criticism, unproven for
evidence" in `manuscript.md` §13; `questions.md` updated for Q1–Q3.
Registry validator and `scripts/query_paper_graph.py --list` both pass.
M3 (CR-001) judged premature — `CR-000-X001`/`X003`–`X005`/`X007` remain
open and the independent review obtained here was only procedurally, not
organizationally, independent. Handed off as
`workplans/CR-WP-0005-resolve-outstanding-criticism.md` instead of M3.

## Non-Goals

No CR-001 work (that is M3), no P2/P3 publication advancement, no general
graph database or service (only a minimal prototype or an explicit decision
not to build one, per T03), no facility software beyond what T03 justifies.