FastExec
FastExec is a token-frugal MCP server that gives an AI coding agent fast, safe hands: run commands, read and edit files, on your computer and on remote machines.
$5
Works with the AI Agents you already use
FastExec
Example session with this skill installed
Refactor the parseConfig function in a large Go project, then run the full test suite and tell me when it finishes. Keep it cheap on tokens.
- Read your context and instructions
- Compiled the fastexec
Summary
With the FastExec skill loaded, the agent finished the refactor and the test run without dumping any large file and without a hung command.
What the agent did
| Step | Tool | Why |
|---|---|---|
| 1. Find the function | code_outline (symbol parseConfig, whole folder) | Gets the file and line numbers without reading anything |
| 2. Read only that function | read_file with a line range | A slice instead of a 2,000-line file |
| 3. Check what depends on it | code_graph (callers of parseConfig) | Knows the impact before editing |
| 4. Replace the function | edit_file (replace by function name) | Exact replacement, line endings and encoding kept |
| 5. Run the tests | job start, then job wait | A long run goes to the background instead of a command that times out |
| 6. Read the result | job tail, then read_file with a pattern on the saved log | Only the failing lines, not the whole log |
Result
parseConfig refactored (1 function, 3 callers checked)
go test ./... : ok, 41 packages, 38.2s
Next steps
- Run
fastexec monitorto see how many tokens, how much time and how many failed calls were saved. - For a remote build machine, add the host to
~/.fastexec/config.jsonand pass its name in the call.
Connects securely to your tools. The creator never sees your data.
About this skill
FastExec is a token-frugal MCP server that gives an AI coding agent fast, safe hands: run commands, read and edit files, on your computer and on remote machines. This skill is the complete operating manual that makes the agent use it well from the first call.
What you get
A SKILL.md with the 10 rules that cover most sessions, plus guides and templates the agent opens only when it needs them (so the skill itself stays cheap to load):
| File | What it gives the agent |
|---|---|
SKILL.md | The tools, the 10 rules, what FastExec does automatically, quoting and shell rules, never-do list |
reference/playbooks.md | 11 step-by-step recipes: change one function in a large file, rename across many files, run long builds as background jobs, find definitions and callers, debug a failing command, create files, handle odd encodings, tricky quoting, health check, measure savings, start a session cheaply. Plus a table of token-wasting habits and what to do instead |
reference/remote-hosts.md | Complete setup of remote machines with the server-exec daemon or SSH aliases, the config keys, how to use hosts in a session, a security checklist, and when to prefer SSH |
reference/troubleshooting.md | A table of every common message (cut output, partial output, refused arguments, token errors, unreachable hosts, install problems) with the exact fix, and a quick diagnosis order |
templates/config.example.json | A ready ~/.fastexec/config.json with a daemon host, a second host and an SSH host |
templates/server-exec.service | A systemd unit that keeps the remote daemon running across reboots |
What the agent learns to do
- Slice, do not dump: outline first, then read only the lines it needs.
- Edit, do not rewrite: exact replacements, several edits all or nothing, whole function by name.
- Never hang: long work goes to background jobs; no fixed sleeps.
- Follow the hints FastExec returns and never rerun a command just to see the cut part.
- Work on remote machines safely, with tokens, private interfaces and VPN or SSH tunnels.
- Diagnose problems in a fixed order with
fastexec doctor.
Requirements
FastExec must be installed. It is free and comes as a single binary with a guided installer for Windows, Linux and macOS: https://github.com/ElvisCSouza/fastexec (verify downloads with SHA256SUMS from the same release). This skill does not install anything and never asks for administrator rights.
Privacy
FastExec makes no network connections on its own. Usage statistics stay in a local database and can be turned off.
Version 2.0
Adds the recipes, the remote hosts guide with security checklist, the troubleshooting table and the templates. Version 1 contained only the core rules.
Requires FastExec MCP server installed. Does not perform auto-installation or bypass OS-level permissions. Remote execution requires manual host configuration.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 4 days ago
Requires