BB digital-twin
Build a psychologically grounded Digital Twin personality skill that makes an agent speak, think, decide, and adapt like a specific real person. Use this skill whenever the user asks to create a digital twin, personality clone, shadow persona, AI stand-in, or personality skill for a named person — including phrasings like "make an AI version of [name]", "clone [name]'s personality", "build a persona for [name]", "create a shadow skill for [name]", "train a twin on [name]", or "I want the agent to respond as [name]". Also use it when the user asks to update or refresh an existing personality skill with newer data. The twin is built by analyzing samples of the target person's own communication across four psychological and linguistic pillars; the output is an installable {name}_personality skill that matches their speech patterns, thinking style, decision-making, and audience-awareness. IMPORTANT: this skill does NOT connect to any data source itself and does NOT require, request, or store any API keys, tokens, or credentials. It sources its training data entirely from connections the user has ALREADY set up — any meeting/call transcript service (Fireflies, Otter, Fathom, Granola, Zoom, Teams, etc.), email, Slack, Teams chat, document stores, or other MCP connectors/skills the user controls. It is a consumer of whatever the user has connected, not an integration. This skill builds personality, voice, and judgment — not factual memory or recall. Pair it with a vector database for memory if full digital twin fidelity is needed.
As a process B 72/100 · Nearly there — weak spots: result and completion
How to improve
- Shorten the description to 1024 characters.
- 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 1550 chars, limit 1024 - note
description-budgetdescription takes 1550 of the ~15000-char shared budget for all skills
Process rating: all ten parameters 72/100
- 0Result and completion. Does not say what the result is
- 70When it triggers. States when to use, but not when not to
- 70Inputs and preconditions. Inputs and preconditions are listed
- 70Failures and branches. 8 branches
- 70Execution cost. Instruction body is 4113 tokens
- 85Steps. 79 steps, 1 vague phrases
- 100Tools and files. No external tools needed
- 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
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
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1549: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +5Description quotes 6 example trigger phrases
- +4Structure: 15 headings
- +3Step-by-step instructions: 79 items
- +4Has examples (1 code blocks)
- +4Reference files are cited in the instructions (5 of 5)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 62.