Products, pilots,
and experiments
AFL’s own products, private pilots, and experiments across cloud infrastructure and AI workflows. These are product explorations; each linked page describes its scope and availability.
Try a skill on
a task you know
AFL skills are reusable instructions and reference material for coding agents. Try the free IAM skill, review the output, and tell me where it helped or fell short.
IAM Least Privilege
A free skill to help an agent draft and review AWS IAM policies. It includes guidance for scoping actions, resources, and conditions. Validate the policy against your workload before deploying it.
Get the free IAM skill →Guardrails & Discipline packs
Paid collections for infrastructure review and coding workflows. Review the included skills, installation instructions, and individual or team licenses before choosing a pack.
Compare the paid packs →Projects
AFL product work and experiments. Follow a project link for its current scope and availability; these descriptions are not customer case studies.
15 projects
Paritai
Claim review and appeal preparation for mental health practices. The founding-practice program seeks workflow feedback; billing and legal decisions require human review.
Technologies
Agora Router
Coordinating multiple autonomous agents without each knowing others' internals; provides a single entrypoint for cross-agent orchestration.
Technologies
Akesis
Explores cloud resource discovery, cost analysis, and recommendations through an API and WebSocket workflow.
Technologies
Anastasis
Explores denial-letter parsing and appeal drafts for therapist billing workflows. Drafts and legal interpretations need qualified review.
Technologies
Athena
Explores lesson generation, active recall, and spaced repetition, with generated material available for review.
Technologies
HE/RM/ES
Explores market research and drafts product briefs, build plans, and experiments to test demand with prospective customers.
Technologies
efix
Investigates reported bugs, gathers reproduction evidence, and prepares proposed fixes and tests in pull requests for human review.
Technologies
Metron
Explores model routing, caching, and usage reports to help teams compare AI cost and response quality.
Technologies
MYTH/OS
Generates story and illustration drafts through a revision workflow. Authors review and edit the results.
Technologies
N.A.R.R.A.T.E. + V.I.S.U.A.L.
Explores a workflow for drafting narratives and accompanying images from a shared brief, with editorial review.
Technologies
Pre-Pared
AI-powered meal planning with auth, billing, and AI recipe generation deployed across Auth0, Stripe, and GCP.
Technologies
Schola
Explores tutoring workflows for placement, spaced review, and learning summaries that tutors can inspect.
Technologies
SO/LA/CE
An experiment in local reflection and wellbeing prompts. It is not a clinical service and has no demonstrated treatment outcomes.
Technologies
Nexus
Routes incoming tasks to the right execution path (fast single agent vs multi-agent pipeline) with shared contracts and observability.
Technologies
LatchSpec
Tests recorded agent tool actions against rules for approvals, ordering, arguments, and limits. Results cover the scenarios tested.
Technologies
Paritai explores claim review and appeal preparation for mental health practices. Its founding-practice program invites feedback on the workflow; billing and legal decisions require qualified human review.