agentsope/SkillAlchemy / skills/LEAP
LEAP
VerifiedLEAP builds skills through two pipelines: Branch A distills a skill from raw data, while Branch B combines multiple skills into one. It is called by the main SkillAlchemy workflow. Use when SkillAlchemy requires distillation or fusion.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 30.5 KB |
| domains/cybersecurity/domain.md | markdown | 2.1 KB |
| domains/energy/domain.md | markdown | 2.1 KB |
| domains/finance/domain.md | markdown | 1.9 KB |
| domains/healthcare/domain.md | markdown | 2.1 KB |
| domains/manufacturing/domain.md | markdown | 2.1 KB |
| domains/mathematics/domain.md | markdown | 2.1 KB |
| domains/media-content-production/domain.md | markdown | 2.4 KB |
| domains/natural-science/domain.md | markdown | 2.0 KB |
| domains/office-white-collar/domain.md | markdown | 2.1 KB |
| domains/persona-os/domain.md | markdown | 5.3 KB |
| domains/robotics/domain.md | markdown | 1.9 KB |
| domains/software-engineering/domain.md | markdown | 2.0 KB |
| references/skill-grammar.md | markdown | 17.3 KB |
| scripts/build_component_index.py | script | 23.7 KB |
| scripts/build_corpus.py | script | 11.0 KB |
| scripts/download_subtitles.sh | script | 2.1 KB |
| scripts/score_skill.py | script | 6.8 KB |
| scripts/srt_to_transcript.py | script | 3.3 KB |
| skill.json | config | 132 B |
domains/persona-os/domain.md
View on GitHub# Persona OS Domain Pack
Purpose:
Extract a person's operating system — how they make decisions, process failure,
handle disagreement, update beliefs, and allocate attention. Used as a
cross-cutting layer when `skill_mode: "persona"`.
This pack is layered ON TOP of the primary domain pack. Read both. The primary
domain pack tells you WHAT the person works on; this pack tells you HOW they
operate.
Common sources:
transcript_interview, longform_text, secondary_criticism, video_subtitle,
social_media, audio_podcast.
Candidate operation families:
decide_under_pressure, recover_from_failure, resolve_value_conflict,
handle_disagreement, update_beliefs, allocate_attention, communicate_decision,
build_team, navigate_power, maintain_consistency.
Validation focus:
Every operation must be traceable to a specific verifiable incident, not just
a quote. The person saying "I believe in X" is low confidence. The person
choosing X when it cost them Y is high confidence.
Research dimensions (for subagent planning — source modality × evidence depth):
- **decision_under_constraint**: How they make choices when resources (time,
money, people, reputation) are tight. What is their tiebreaker? Do they
optimize for speed, quality, principle, or survival? Look for moments where
they had to choose between two goods or two evils.
*Revealed by:* interviews about crisis moments, biographies, post-mortem
analyses, their own retrospective accounts.
- **failure_processing**: How they metabolize failure. Do they reframe it as
learning? Externalize blame? Extract specific rules? Change behavior
measurably? The most important signal: compare their rhetoric BEFORE and
AFTER a major failure. Did they actually change what they do, or just
how they talk about it?
*Revealed by:* long-timespan interviews (pre- and post-failure), secondary
analysis tracking their actions over time, their own post-mortem accounts.
- **value_conflict_resolution**: When two stated values collide, which wins?
Everyone says they value both "quality" and "speed." When forced to trade
one for the other, which do they actually sacrifice? This reveals their
true value hierarchy, not their stated one.
*Revealed by:* product decisions, hiring/firing choices, deal negotiations,
public controversies where they had to pick a side.
- **attention_allocation**: What do they pay attention to and what do they
deliberately ignore? How do they decide what NOT to do? This is often more
revealing than what they DO — ignored opportunities are invisible in
retrospect but define their trajectory.
*Revealed by:* interviews where they discuss what they said no to, product
roadmaps showing features they killed, strategic pivots where they abandoned
a direction.
- **influence_and_disagreement**: How they handle being challenged. Persuade
with logic? Override with authority? Delegate and step back? Avoid the
conflict? Their response to disagreement reveals their theory of how truth
is reached — through debate, through authority, through data, through
intuition?
*Revealed by:* meeting accounts, co-founder dynamics, public debates,
their own descriptions of internal disagreements.
- **learning_and_updating**: How they change their mind. What kind of evidence
makes them update? A personal experience? Data? Someone they trust telling
them? Public pressure? How long does it take? The most valuable extraction:
identify a belief they once held strongly and later reversed. What caused
the reversal?
*Revealed by:* before/after comparisons on specific topics, interviews
where they say "I used to think X, but now I think Y," actions that
contradict their earlier stated positions.
- **identity_and_narrative**: What story do they tell about themselves? How
does this narrative constrain or enable their actions? The "I'm the kind of
person who..." frame is often the deepest OS layer — it determines which
options even appear on their menu.
*Revealed by:* autobiographical accounts, brand/personal mythology, how
they introduce themselves, the metaphors they use to describe their role.
Cross-domain bridges (how persona dimensions interact with primary domains):
- software-engineering: decision_under_constraint maps to architectural
tradeoffs; attention_allocation maps to feature prioritization
- office-white-collar: influence_and_disagreement maps to organizational
politics; value_conflict_resolution maps to stakeholder tradeoffs
- media-content-production: identity_and_narrative maps to personal brand
construction; learning_and_updating maps to content iteration
- finance: decision_under_constraint maps to risk management; failure_processing
maps to loss recovery
- Any domain: persona dimensions are domain-agnostic. The primary domain
provides the SCENARIOS; persona-os provides the EXTRACTION LENS.
Dimension derivation hints (when source evidence suggests a dimension not on the menu):
- source reveals an unusual relationship with time (extreme patience or
impulsivity) → derive `temporal_orientation` dimension
- source reveals a distinctive relationship with money (unusual frugality or
spend patterns) → derive `resource_philosophy` dimension
- source reveals a mentorship or apprenticeship pattern → derive
`knowledge_transfer` dimension
- new dimension rule: name = `<operation_the_source_exposes>`, mission =
"extract how <target> performs <operation>"
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |