You embed at a client site. You bridge strategy and code. You ship on their systems, not yours.
Every morning you open your AI coding agent, and it has no idea what happened yesterday. You re-paste the same context. You explain the stakeholders again. You remind it about the scope change from Tuesday. Meanwhile, the real problem - the one the brief didn't mention - sits undiscovered because nobody asked the right questions on day one.
fdeops fixes this. It gives your AI agent a complete engagement methodology and a private memory that writes itself. You type @fde, describe your situation, and the right method runs - from first stakeholder meeting to final handoff. Tomorrow's session starts exactly where today ended.
flowchart LR
A["@fde"] --> B{"Describe\nyour situation"}
B --> C["Embed & Trust"]
B --> D["Discover & Diagnose"]
B --> E["Plan & Align"]
B --> F["Build & Guard"]
B --> G["Ship & Verify"]
B --> H["Operate & Close"]
C --> I[".fde/ memory\n(written as you work)"]
D --> I
E --> I
F --> I
G --> I
H --> I
I --> J["Next session\nloads automatically"]
Works with Claude Code · Cursor · Copilot · Devin · Gemini CLI · Ollama · LM Studio · any model that reads SKILL.md
The CLI
The Fieldbook Dashboard
| You are... | fdeops helps when... |
|---|---|
| Consultant or contractor at a client site | Every session, you re-explain context. fdeops remembers for you. |
| Solutions architect bridging strategy and code | You navigate politics AND architecture. fdeops has methods for both. |
| Agency engineer running 3-5 clients | Client details blur together. One .fde/ per customer, never cross-contaminated. |
| Forward Deployed Engineer | The role this was built for. 34 skills across the full engagement lifecycle. |
| Technical founder doing client work solo | You ARE the team. The agent becomes your second brain. |
| Enterprise programme lead | Leading AI transformations? Built-in methods for model selection, agent safety, governance, and cost management. |
| Without fdeops | With fdeops | |
|---|---|---|
| Monday morning | Re-paste last week's context, explain the stakeholders again | Agent opens with "last session you were on the ingest retry - CTO demo is Friday" |
| Scope creep | Five "small" additions absorbed silently, timeline slips | Receipts timestamped - you walk into the sponsor meeting with evidence |
| Multiple customers | Wrong client name in a status update, details blur | One folder per customer, context-switch protocol, cross-contamination checklist |
| The sponsor meeting | "We completed the API endpoint" | "Manual reconciliation dropped from 3 FTEs to 0.5 - here's the rollback if it turns" |
Requires: Node.js >= 18
npx fdeops init my-client # creates engagement at ~/fde-engagements/my-client/.fde/
export FDEOPS_ENGAGEMENT=~/fde-engagements/my-client/.fdeThen in your AI agent:
@fde I'm on site. First stakeholder meeting tomorrow. Brief says fix the payments API.
Day 1 - it coaches your preparation:
Brief says payments API - but nobody named who owns budget. That's
your first question. What does failure look like to the sponsor
personally? I'll draft brief.md and stakeholders.md after you come
back with notes.
Day 15 - it picks up where yesterday ended:
Last session: ingest retry on staging. CTO demo Friday, status hasn't
gone out since Tuesday. Globex quiet for 3 days - trust signal amber.
Send CTO status first, then finish retry, then check Globex.
The memory compounds. You maintain nothing.
Other install methods:
npx skills add suboss87/fdeopsorgit clone && node bin/install.js
One skill file powers every tool. Install adapters for your setup:
npx fdeops adapters ~/fde-engagements/my-client| Tool | What it reads |
|---|---|
| Claude Code | Plugin + ~/.claude/FDEOPS-CLAUDE.md |
| Codex / OpenAI / generic | AGENTS.md |
| Gemini CLI | GEMINI.md |
| Cursor | .cursor/rules/fde.mdc |
| GitHub Copilot | .github/copilot-instructions.md |
| Local LLMs (Ollama, LM Studio, llama.cpp, vLLM) | Load SKILL.md as system prompt |
Each adapter points at the same @fde skill, so the methodology and memory stay consistent across tools. Details: adapters/.
No cloud dependency. fdeops calls no external API. The AI skill is a markdown file your model reads. The CLI is local Node.js. Works fully offline, fully air-gapped, fully private. See
adapters/LOCAL-LLM.mdfor local model setup.
YOU (human) AI CODING AGENT (software)
meetings, judgment @fde routes -> right skill -> drafts the artifact
| -----> .fde/ memory (written as you work)
| |
+---------------> client workspace (code, VPN, tickets)
- Describe - tell the agent what's happening ("new client", "production is down", "need a board update")
- Route - the system picks the right skill from 34 options across 6 domains
- Execute - the skill's method runs, artifacts are written to
.fde/, you review at checkpoints
| Domain | Skills | What it covers |
|---|---|---|
| Embed & Trust | land, audit, stakeholder-radar, trust-engineering, scope-defense | First days: access, credibility, scope |
| Discover & Diagnose | discover, assumption-audit, use-case-scoring, sketch | Finding the real problem behind the brief |
| Plan & Align | plan, business-case, options-analysis, initiative-triage | Sequencing work, getting sponsor alignment |
| Build & Guard | build, incremental-build, test-on-legacy, blast-radius, debug, rescue, security-audit, observability | Helping you build safely on their codebase |
| Ship & Verify | ship, review, rollback-drill, qa-live | Getting to production without surprises |
| Operate & Close | status, demo-prep, debrief, exec-narrative, dashboard, multi-customer-ops, close, handoff-engineering, pattern-extract | Running and ending the engagement well |
Overlays activate automatically when your engagement involves AI projects, executive reporting, fintech, healthcare, or government compliance.
Full skill details: docs/skills-reference.md
Your fieldbook - one per client, private to you, plain markdown:
| File | Role | Written by |
|---|---|---|
context.md |
Where you are; loaded first every session | every phase + session-stop hook |
brief.md |
What they said - hypothesis until discover | land |
success.md |
Done, measured, signed-off by whom | land |
reality.md |
The real problem, with evidence | discover / audit |
terrain.md |
Codebase map: hotspots, test gaps, AI components, data estate | discover / audit |
stakeholders.md |
Champions, resistance, trust signals | land, updated continuously |
trust-profile.md |
Sacred data, AI policy, approval chain | land + overlays |
decisions.md |
Plan + choices + integration contracts + sizing | plan / build / review / rescue |
risks.md |
Live risk register | all phases |
delivery.md |
What shipped, business value, rollback, pulse, adoption metrics | build / ship |
Every claim is tagged with its source and date so you can defend it in front of skeptical stakeholders.
These commands run locally on your machine. No AI needed, no API costs, works offline.
fde scan # day-1 recon: hotspots, test gaps, "temporary" code, AI components, secrets
fde resume # initialize or resume an engagement
fde log # write decisions, risks, delivery, contacts
fde receipts # search memory with dates
fde capture # session-end snapshot
fde status # portfolio triage across all customers
fde dashboard # render every engagement into one offline dashboard- The artifact is the memory - producing work and recording it are one action
- Trust before production - earn the right to touch their systems
- Brief is a hypothesis - discover before building the wrong thing
- Evidence on every claim - these files get defended in front of skeptical clients
- Map before moving - unknown terrain gets characterisation tests
- Thin slices - ship learning, not theatre
- One customer, one folder - context never bleeds
cd fdeops && git pull && node bin/install.jsMaintained by Subash Natarajan. Feedback via Issues - see CONTRIBUTING.md.
FDE Methodology · ATTRIBUTION.md · SECURITY.md · PRIVACY.md · Repo layout · Skills reference · MIT