tegwick
3aa8cb7eaf
The channel publishes a post written from a hall entry -- condensed, personal, reach-optimized -- not the entry itself. That changes what this interface is for and where its boundary sits. Composition is deliberately outside the interface. Composing is a probabilistic editorial task, and Blueprint 48.1 forbids an LLM in the mandatory request path, so an interface that composed on demand would be non-deterministic exactly where determinism is the whole claim. It is also editorial: a post speaks about a named person's work in HelixForge's voice, and a review step inside the request path is a review step nobody performs. The contract now takes a composed post with a required reviewer, a cited source entry and a consent basis. A post carrying none of those is refused. Absence of objection is not consent, and in an audit trail an unfaithful post looks exactly like a faithful one -- the defence is that a human read it, the source is named, and both are recorded. The first experiment is rebuilt accordingly. The old competition asked how to fit a whole entry into Telegram; that question no longer exists. The new one asks whether the portrait earns the 3072 characters of caption limit it costs, which is a delivery question this interface actually owns. Competition over editorial voice belongs to the campaign and runs on the per-variant engagement this interface reports. Adds T05b (composition and review), T05c (consent basis for the 94 existing participants, blocking public publication) and T11 (extracting the campaign into pr-hall-of-helix). Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014KmVxhJ35tCo7rE7UnLwWu Assistant: claude-code Assistant-Model: opus Assistant-Process: 1116572@bnt-lap001 Assistant-Session: 8ba9bb93-a72a-4883-b189-2499cce5c400
68 lines
1.9 KiB
YAML
68 lines
1.9 KiB
YAML
# The interface's first experiment: does the portrait earn its place?
|
|
#
|
|
# This tests DELIVERY, which this interface owns. The competition about which
|
|
# editorial voice travels belongs to the campaign and is run on the per-variant
|
|
# engagement this interface reports — see docs/observation.md.
|
|
#
|
|
# fluid experiment design --file experiments/E-tg-delivery.yaml
|
|
# fluid experiment start E-tg-delivery --generation 2 \
|
|
# --default-revision R-2 --policy-out rp.json
|
|
# fluid policy put --file rp.json
|
|
fluid_experiment:
|
|
schema_version: "0.1"
|
|
|
|
id: "E-tg-delivery"
|
|
interface_id: "helix-forge-telegram-publishing"
|
|
|
|
hypothesis_refs:
|
|
- "H-tg-visual"
|
|
- "H-tg-textonly"
|
|
|
|
# R-2 attaches the portrait, R-3 publishes text only. Neither is an
|
|
# incumbent: R-1 publishes without either treatment being settled, so this
|
|
# is a comparison between two answers rather than against a baseline.
|
|
control_revision: "R-2"
|
|
candidate_revisions:
|
|
- "R-3"
|
|
|
|
cohorts:
|
|
- "telegram-subscribers"
|
|
|
|
allocation:
|
|
control: 0.5
|
|
candidate: 0.5
|
|
|
|
# An even split because there is no incumbent to protect. The usual 90/10
|
|
# caution limits exposure to an unproven candidate; here both arms are
|
|
# equally unproven and the scarce resource is posts, not safety.
|
|
|
|
metrics:
|
|
primary:
|
|
- "engagement_rate"
|
|
guardrails:
|
|
- "unreviewed_publication"
|
|
- "missing_source_attribution"
|
|
- "missing_subject_consent"
|
|
- "error_rate"
|
|
secondary:
|
|
- "source_link_rate"
|
|
- "forward_rate"
|
|
learning:
|
|
- "time_to_first_view"
|
|
|
|
start_conditions:
|
|
- "both revisions verified against the private test channel"
|
|
- "channel has at least 100 subscribers"
|
|
|
|
stop_conditions:
|
|
- "hard_guardrail_violation"
|
|
- "unreviewed_publication"
|
|
- "manual_stop"
|
|
- "max_duration_reached"
|
|
|
|
max_duration_hours: 2160 # 90 days
|
|
|
|
result:
|
|
state: "PLANNED"
|
|
preferred_revision: null
|
|
evidence_refs: []
|