David Duncan
Community@davdunc · Oaxaca, MX
I am ex-AWS and ex-Red Hat. I am a contributor to open source operating systems and I function as a public cloud Solutions Architect.
Agent Skills by David Duncan
Showing 57 vetted skills indexed across 2 GitHub repositories.
PAI
Orchestrate Claude Code through a 7-phase loop with verifiable Ideal State Criteria.
Telos
Organize personal and project context into TELOS life files and generate narrative and consulting-style reports.
Research
Produce verified structured research outputs across multi-agent investigation workflows.
RootCauseAnalysis
Convert incident reports into blameless root-cause analyses with RCA workflows.
ApertureOscillation
Oscillate question scope across tactical and strategic passes to identify design tensions.
Evals
Run regression and capability evaluations for multi-turn AI agent workflows.
WorldThreatModel
Stress-test ideas against 11 world threat horizons from 6 months to 50 years.
Optimize
Iteratively modify targets and select improvements using numeric metrics or LLM-as-judge evals.
CreateCLI
Generate Bun-based TypeScript CLI projects with strict typing and JSON output.
PAIUpgrade
Generates prioritized upgrade recommendations for PAI based on audits, TELOS analysis, ecosystem signals, and algorithm reflections.
ISA
Scaffold, interview, audit, reconcile, or seed Ideal State Artifacts with verification evidence.
ExtractWisdom
Extract content-adaptive wisdom from YouTube transcripts, articles, and pasted text into structured markdown.
Council
Generate multi-perspective debate transcripts with round-based expert deliberation.
Sales
Convert product documentation into sales narratives, charcoal sketches, and talking points.
WriteStory
Build a PRD-style Story Bible and draft chapter prose across seven narrative layers.
BitterPillEngineering
Audit AI assistant instruction sets and classify over-prompting rules into categorized outputs.
Security
Route security assessments across recon, web testing, and prompt-injection scenarios.
RedTeam
Decompose claims and run adversarial critique from 32 parallel experts.
AudioEditor
Generate word-level transcripts, classify segments, and apply ffmpeg edits with crossfades to clean audio recordings.
PrivateInvestigator
Locate individuals and verify identity via public-data OSINT with confidence scoring.
Science
Convert goals, hypotheses, and experiments into an evidence-based decision loop.
ContextSearch
Search session registries, MEMORY work directories, and ISA titles for topic keywords.
Delegation
Routes multi-workstream tasks to specialized agents or Claude Code Teams for parallel execution.
Utilities
Automate developer tooling workflows including CLI generation, skill scaffolding, and evaluations.