fluid-core/workplans/FLUID-WP-0009-campaign-repos-and-example-separation.md

id	type	title	domain	repo	status	flavor	owner	topic_slug	created	updated	planning_priority	planning_order	depends_on	related_repos	state_hub_workstream_id
FLUID-WP-0009	workplan	Establish the pr- campaign repository pattern and reduce hall-of-helix to an example	infotech	fluid-core	blocked	implementation	worsch	fluid-core	2026-09-04	2026-09-27	medium	9	FLUID-WP-0008	fluid-telegram fluid-x fluid-substack hall-of-helix helix-forge	b6459dd0-fc4f-54a4-a2e9-d7b83cff6c6e

FLUID-WP-0009 — Campaign repositories and example separation

Two things this framework needs before it is genuinely reusable, and neither is urgent enough to block the first Telegram publication.

A place for campaigns to live. Publishing the Hall of Helix is an ongoing public-relations offering with its own editorial voice, cadence, consent records and budget. Today it lives inside FT-WP-0001, a workplan that is supposed to finish. It needs a repository.

A framework that is not about one campaign. While the only worked example is the Hall of Helix, it is hard to tell which parts of fluid-core and fluid-telegram are general and which quietly assume that particular use. The test of a framework is a second consumer, and the honest way to find the accidental couplings is to look for them deliberately.

This workplan should start after FT-WP-0001 T12 — the loop closing once on real traffic. Extracting a seam before using it is how you learn you cut it in the wrong place.

T01 — Define the pr- repository convention

id: FLUID-WP-0009-T01
status: done
priority: medium
state_hub_task_id: "2a3aab85-e464-5b17-980f-1896ecd3794f"

Write down what a pr- repository is, so the second one does not have to be reverse-engineered from the first.

The prefix covers public relations, promotion and propaganda — deliberately including the uncomfortable third word, because a campaign repository that cannot name what it is doing will not govern it well.

A pr- repo should carry: its own INTENT.md; a campaign intent describing audience, voice and what it will not say; consent records for anyone it writes about; composition strategies as named variants; and a publishing cadence. It consumes one or more fluid-* interfaces and owns none of them.

Written: docs/campaign-repositories.md (2026-09-04.) It adds a review step to the list above, the boundary table against fluid-*, the registration rules (domain and topic follow the subject, since the domains are market sectors and none of them is "marketing"), and the point that a campaign may run its own FLUID loop over its own artifacts without merging it into the interface's.

It states its own test: whether a second campaign can be created from the document without reading pr-hall-of-helix. Until that happens it describes one case rather than a general rule — the same caution T05 applies to the framework.

T02 — Establish pr-hall-of-helix

id: FLUID-WP-0009-T02
status: wait
depends_on: [FT-WP-0001-T12]
priority: medium
blocking_reason: "Campaign move waits until the FLUID loop has closed once (FT-WP-0001 T12); the repository itself already exists."
state_hub_task_id: "7d40da89-9a4d-5358-ae2b-7a1a1339efff"

Create the repository and move the campaign into it, per FT-WP-0001 T13. The split is already written there; this task is the execution and the registration with the State Hub under the infotech domain.

Partly done, ahead of the sequencing above (2026-09-04). The repository exists, is registered (domain infotech, topic helix-forge, prefix PRHOH-WP-), and carries a grounded INTENT.md and SCOPE.md plus the declared Telegram presence at presence/telegram.yaml.

It was created early because provisioning needed somewhere to put the presence spec — a bot's name and a channel's title are editorial, and putting them in the delivery interface would have cut the seam in the wrong place immediately.

What remains is the part the sequencing was actually protecting: the campaign itself has not moved. Composition and the review queue still live in fluid-telegram under FT-WP-0001 T06/T07, and should stay there until the loop has closed once (T12), as FT-WP-0001 T13 says.

The campaign becomes a consumer of fluid-telegram in cohort hall-publishing-jobs, and runs its own FLUID loop over its own artifacts — which composition variant travels — using the per-variant engagement the interface reports.

T03 — Find what fluid-core assumes about the Hall of Helix

id: FLUID-WP-0009-T03
status: done
priority: medium
state_hub_task_id: "d9c69244-19ef-5eaf-9f43-d0374d6278de"

Audit fluid-core for anything that assumes this particular campaign. The conformance handover test already couples the two repositories deliberately; check whether anything else does so by accident.

Known candidates to check rather than assume clean: default redaction patterns, the classifier's thresholds, the fitness metric names, and whether examples/echo-interface is doing enough work to be the general example it claims to be.

Audited (2026-09-27). No accidental coupling found:

• internal/observation/redact.go DefaultRedactionPolicy — patterns target bearer tokens, emails and connection strings; nothing Telegram- or hall-of-helix-specific.
• internal/observation/classify.go DefaultClassifierOptions — thresholds (RepeatThreshold, MinConsumers, ErrorRateThreshold) are generic interaction-shape numbers, not tuned to one interface's traffic.
• internal/fitness/fitness.go — metric names are caller-supplied strings (MetricSpec.Name); the package carries none of its own.
• examples/echo-interface — already the general §33 worked example (GET /v1/entries / GET /v1/entries/latest); it has no dependency on the Telegram case and is deliberately kept separate from it.
• One accidental reference: a comment in internal/runtime/connector.go naming "hall-of-helix" as the illustrative example for why non-idempotent methods aren't retried. Generalized to "the same resource" — no functional coupling, just a stray name in prose.
• The only deliberate coupling is the conformance handover test (conformance/loop/loop_test.go, internal/control/api_test.go) using hall-publishing as its interface ID, exactly as this task already expected.

Conclusion: fluid-core does not quietly assume the Hall of Helix campaign. T05's second-interface test remains the real proof, but nothing found here would block it.

T04 — Reduce the Telegram case to a documented example

id: FLUID-WP-0009-T04
status: wait
priority: low
depends_on: [FLUID-WP-0009-T02]
blocking_reason: "Campaign must live elsewhere first (T02)."
state_hub_task_id: "ec77e62c-1251-5331-a1a8-42985fb97b94"

Once the campaign lives elsewhere, docs/handover/fluid-telegram.md becomes a worked example rather than a live handover: here is a real interface, here is the pressure it faced, here is how the loop closed on it.

That is more valuable as documentation than it is as history, and it is the form fluid-x and fluid-substack will want to read.

T05 — Prove reuse with a second interface

id: FLUID-WP-0009-T05
status: todo
priority: low
state_hub_task_id: "227d0d53-f325-5d95-8a18-d56603c820fd"

The real test of T03. Take fluid-x or fluid-substack from empty capability canon to a published R-1, and record honestly how much of the fluid-telegram handover transferred unchanged.

Whatever did not transfer is the accidental coupling, and finding it this way is cheaper than arguing about it in advance.