ACA experiment design
Create a clean experiment with one variable, one metric, and a stopping rule.
Rules
- Do not change multiple variables at once.
- Prefer small tests before scaling.
- Save the experiment plan before applying changes.
- Mutations require approval.
Workflow
- Inspect current assets with
list_campaigns, list_email_sequences, list_lead_lists, list_generation_jobs, and list_autopilots.
- Pick experiment type:
- Audience
- Angle
- CTA
- Channel
- Sequence timing
- Content topic
- Define control, variant, sample size, metric, decision threshold, and review date.
- Save with
create_strategy_document.
- If approved, route to the right executor:
aca-copy-variants
aca-find-leads
aca-launch-outreach
aca-content-week
Output format
Experiment: {name}
Hypothesis: {hypothesis}
Control: {control}
Variant: {variant}
Metric: {metric}
Sample size: {n}
Decision rule: {rule}
Next action: {skill}
Skill chaining
This skill participates in the ACA chain. Preserve the selected ACA org, relevant IDs, user brief, approval state, and any generated artifacts when continuing into another ACA skill. If the user asked for execution and a downstream condition is met, continue into the next skill automatically; otherwise end with the handoff block.
Upstream
- Called by
aca-weekly-rhythm, aca-positive-reply-scoring, aca-copy-variants, or optimization requests.
Auto-continue conditions
- Copy variable -> continue to
aca-copy-variants.
- Audience variable -> continue to
aca-find-leads or aca-lead-quality.
- Content variable -> continue to
aca-content-week.
- Launch-ready test -> continue to
aca-launch-outreach.
Stop before chaining when
- Ask before applying experiment changes or launching tests.
Downstream skills
aca-copy-variants - create message variants.
aca-find-leads - build audience variant.
aca-lead-quality - segment audience variant.
aca-content-week - create content variant.
aca-launch-outreach - launch approved experiment.
Handoff block
Chain state: {continue|needs_approval|blocked|complete}
Next skill: {aca-skill-name|none}
Reason: {why this handoff is or is not needed}
Carry forward: {org_id/name, product_id, icp_id, lead_list_id, campaign_id, sequence_id, job_id, approvals, constraints}
ACA tools used
list_campaigns, list_email_sequences, list_lead_lists
list_generation_jobs, list_autopilots
create_strategy_document
1---2name: aca-experiment-design3description: Design one controlled ACA experiment for outbound or content. Use when the user wants to improve reply rate, test an angle, split audiences, A/B test copy, compare channels, or plan experiments.4license: MIT5---67# ACA experiment design89Create a clean experiment with one variable, one metric, and a stopping rule.1011## Rules1213- Do not change multiple variables at once.14- Prefer small tests before scaling.15- Save the experiment plan before applying changes.16- Mutations require approval.1718## Workflow19201. Inspect current assets with `list_campaigns`, `list_email_sequences`, `list_lead_lists`, `list_generation_jobs`, and `list_autopilots`.212. Pick experiment type:22 - Audience23 - Angle24 - CTA25 - Channel26 - Sequence timing27 - Content topic283. Define control, variant, sample size, metric, decision threshold, and review date.294. Save with `create_strategy_document`.305. If approved, route to the right executor:31 - `aca-copy-variants`32 - `aca-find-leads`33 - `aca-launch-outreach`34 - `aca-content-week`3536## Output format3738```text39Experiment: {name}40Hypothesis: {hypothesis}41Control: {control}42Variant: {variant}43Metric: {metric}44Sample size: {n}45Decision rule: {rule}46Next action: {skill}47```4849## Skill chaining5051This skill participates in the ACA chain. Preserve the selected ACA org, relevant IDs, user brief, approval state, and any generated artifacts when continuing into another ACA skill. If the user asked for execution and a downstream condition is met, continue into the next skill automatically; otherwise end with the handoff block.5253**Upstream**54- Called by `aca-weekly-rhythm`, `aca-positive-reply-scoring`, `aca-copy-variants`, or optimization requests.5556**Auto-continue conditions**57- Copy variable -> continue to `aca-copy-variants`.58- Audience variable -> continue to `aca-find-leads` or `aca-lead-quality`.59- Content variable -> continue to `aca-content-week`.60- Launch-ready test -> continue to `aca-launch-outreach`.6162**Stop before chaining when**63- Ask before applying experiment changes or launching tests.6465**Downstream skills**66- `aca-copy-variants` - create message variants.67- `aca-find-leads` - build audience variant.68- `aca-lead-quality` - segment audience variant.69- `aca-content-week` - create content variant.70- `aca-launch-outreach` - launch approved experiment.7172**Handoff block**7374```text75Chain state: {continue|needs_approval|blocked|complete}76Next skill: {aca-skill-name|none}77Reason: {why this handoff is or is not needed}78Carry forward: {org_id/name, product_id, icp_id, lead_list_id, campaign_id, sequence_id, job_id, approvals, constraints}79```8081## ACA tools used8283- `list_campaigns`, `list_email_sequences`, `list_lead_lists`84- `list_generation_jobs`, `list_autopilots`85- `create_strategy_document`