| name | me |
| description | Plain reference facts about the user's machines and self-hosted services, and which to prefer for a given task. No secrets, hostnames, or IPs live here — load for context, then ask the user for current endpoints or credentials. Use whenever a task involves the user's own infrastructure: deploying or self-hosting something, using a local LLM or GPU, or asking what machines, servers, endpoints, or services they have. Triggers on: "my setup", "my infrastructure", "homelab", "my server", "use my GPU", "local model", "where should I deploy", "what services do I have".
|
Me — Personal Context
Facts only, no secrets, no hostnames, no IPs, no file paths. When a task
needs an actual endpoint, API key, password, or token, ask the user directly
in the moment — never store or guess one here.
Identity
- Anish Shrestha, GitHub
anyesh
Machines
- A GPU workstation is available for heavier local inference. Ask the user
for the current host and port before using it.
- A homelab server runs several self-hosted apps (file sync, budgeting,
order management) with scheduled backups.
- Networked storage is available for shared files. Some encrypted folders
need a manual remount after a reboot — ask the user if something looks
unmounted.
Self-hosted services
- A local LLM endpoint is available for lightweight tasks (summarization,
classification, quick generation). Ask the user for the current URL and
key.
- A PII/privacy filtering service is available.
- Self-hosted notification, email, auth, and uptime-monitoring services are
running. Ask the user for current details before relying on any of them.
When a task needs deployment, auth, notifications, email, or monitoring,
check with the user whether a self-hosted option should be preferred over
external SaaS — do not assume specifics.
LLM endpoint preference
- Lightweight tasks (summarization, classification, quick generation):
prefer the user's self-hosted endpoint first. Ask for the current URL and
API key before making a request — never assume either is unchanged.
- Heavier local inference: use the GPU workstation. Ask the user for the
current host and port, then call the OpenAI-compatible endpoint directly.
Use the local LLM proactively for: a second opinion on code or a decision, a
review pass before committing, brainstorming alternatives, or sanity-checking
your own answer.
Updating this file
This is a plain file, not a vault — edit it directly with the Edit tool when
the user shares a new fact. Never add an API key, password, token, IP
address, hostname, domain name, or file path to it — describe capabilities
generically and ask the user for specifics at use time.
