BB outclaw-plan
Build, approve, schedule, monitor, and react to multi-channel B2B outreach campaigns. Reads the user's KB and learned style; constructs multi-step multi-channel plans (messages + social actions); shows previews for per- touchpoint approval; schedules via taskflow; listens for replies and proposes revisions. Also owns campaign dashboard + pause/resume + opt-out handling + anti-spam caps. Any request involving sending, queueing, scheduling, or contacting a person routes here — including bulk-send requests that MUST be refused. Triggers on: 'reach out to <person>', 'contact <person>', 'plan outreach', 'draft cold email', 'draft outreach', 'send cold <channel>', 'follow up', 'show my campaigns', 'campaign status', 'pause <campaign>', 'resume <campaign>', 'cancel <campaign>', 'did anyone reply', 'any responses', 'is anyone gonna reply', 'who replied', 'check replies', 'send it all', 'send them all', 'just send it', 'fire the touchpoints', 'queue N emails', 'send N cold emails', 'send the plan', 'ignore feedback and send', 'override the tone feedback', 'stop contacting me' (when received as a reply from a lead — triggers opt-out flow). Invoked by the outclaw orchestrator.
As a process B 77/100 · Nearly there — weak spots: execution cost
How to improve
- Shorten the description to 1024 characters.
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 7. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1185 chars, limit 1024 - warning
body-longSKILL.md body ≈ 8013 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 77/100
- 40Execution cost. Instruction body is 8013 tokens: crowds the task out of the window
- 60Tools and files. Uses tools (node) that frontmatter does not declare
- 60Result and completion. Output format stated, no completion criterion
- 70When it triggers. States when to use, but not when not to
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Steps. 102 steps
- 100Failures and branches. 8 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- 100Progress reporting. Reports progress
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 11 top-level sections: this looks like several domains in one skill
- low The response is described with custom markup (22 tags): a typed call is more reliable
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1184: 120–800 characters recommended
- -5TODO / placeholder text left in the skill
- +1No license
- +2Single-language instructions
- +4Structure: 27 headings
- +3Step-by-step instructions: 102 items
- +3Output format is stated explicitly
- +4Has examples (15 code blocks)
- +4Reference files are cited in the instructions (1 of 3)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 48.