Atlas Tradecraft logo Atlas Tradecraft Book a free hour
Atlas Tradecraft No. 001 MMXXVI ICM / Interpretable Context Methodology

AI work
your auditors trust.
Your team owns.

Atlas Tradecraft teaches the Interpretable Context Methodology — ICM — a numbered, auditable pipeline that turns one brief into vetted, reviewable AI output, shipped as a portable packet any AI can continue.

A certified specialist embeds one-on-one with your employee, rebuilds their week as an ICM pipeline, and hands back a playbook they keep using.

Claim a free 1-hour session → Try the method at icmproject.com ↗ Read the manifest ↓

Built for leadership at

Fortune 500 firms where AI output has to be reviewable, portable, and defensible — not just produced.

01 The Gap Atlas / ICM · §01

The tools are deployed.
The method isn’t.

Enterprises spend seven figures a year on AI licenses and trainings — yet productivity gains arrive in pockets, not across the org.

A few power users hit 10×. Everyone else stays at 1.0× — and their output isn’t reviewable, repeatable, or portable.

i.
11%

Productivity

of paid AI seats are used weekly

Licenses sit dormant; no method means no momentum.

ii.
3in4

Quality

employees say AI output is inconsistent and unverifiable

Generic training shows buttons, not pipelines.

iii.
< 8%

Auditability

of enterprises can audit how output was produced

Chat logs aren’t an audit trail.

02 The ICM Method Spec v1 · Reference

The folder structure is the agent.

ICM — the Interpretable Context Methodology — is an AI-native framework where the folder structure is the agent. One paragraph in; ICM drafts a numbered pipeline whose stages run in scoped contexts, write reviewable artifacts, and hand off.

A middle stage is a human review checkpoint; the last exports a portable packet — Markdown plus a manifest — that any AI can continue. No lock-in. No black boxes. Auditors can read every step.

  1. 01

    Auto-enrich

    One paragraph becomes a structured brief before a single prompt is written.

  2. 02

    Numbered pipeline

    FRAMER → RESEARCHER → PRODUCER → REVIEWER → SHIPPER. Each stage scoped and reviewable; REVIEWER is the human gate.

  3. 03

    Portable packet

    Export as Markdown + ZIP. Any AI — today’s or 2027’s — keeps the work moving. Your IP, your storage, your auditors.

ICM is Atlas’s curriculum; the tool that implements it lives at icmproject.com. Atlas is the embedded enterprise tier.

03 The Pipeline FRAMER → SHIPPER · 5 stages

One brief in.
A whole pipeline out.

Input One paragraph — the brief
Output packet.json · Markdown + ZIP · any AI can continue
  1. 01↓

    Framer

    Enriches the brief into goals, audience, and voice — then drafts the pipeline.

    01_frame/ · _PROMPT.md
  2. 02↓

    Researcher

    Runs in its own scoped context; writes its own reviewable artifacts.

    02_research/ · findings.md
  3. 03↓

    Producer

    Co-builds the assets, prompts, projects, sub-agents, and automations.

    03_produce/ · draft-vN.md
  4. Human gate 04↓

    Reviewer

    An explicit human checkpoint with a written audit — the record risk and legal ask for.

    04_review/ · audit.md
  5. 05↓

    Shipper

    Exports a portable packet — Markdown plus manifest — any AI can continue.

    05_ship/ · packet.json

04 · REVIEWER — written human checkpoint Every stage writes plain-Markdown, diffable artifacts

04 The Engagement 3 phases · 1 pipeline

Three phases.
One audit-grade pipeline.

Every engagement runs on the ICM pipeline; the stages in monospace are the artifacts your auditors get.

  1. 01

    Frame the work

    ICM · 00 INTAKE · 01 FRAMER

    Your specialist shadows the employee, audits their tools and artifacts, and writes the brief into ICM intake — a structure your auditors recognize from day one.

    • Time-on-task baseline
    • Repetitive surface inventory
    • brief.md · voice.md · _PROMPT.md
  2. 02

    Build & review

    ICM · 02 RESEARCHER · 03 PRODUCER · 04 REVIEWER

    We co-build the assets, prompts, Claude Projects, sub-agents, and automations the employee needs. Each stage writes reviewable output; REVIEWER is a written human checkpoint embedded in the workflow.

    • Custom Claude Projects & sub-agents
    • Agent & automation pipelines
    • findings.md · draft-vN.md · audit.md
  3. 03

    Ship & multiply

    ICM · 05 SHIPPER · packet.json

    We export a portable packet your employee uses Monday and teammates inherit Tuesday. A 90-day roadmap, weekly check-ins, and a productivity dashboard turn one packet into a department’s operating system.

    • Portable packet handoff
    • Weekly throughput review
    • Quarterly executive ROI report
05 Deliverables 6 items · written to your storage

Not a certificate.
A packet you own.

Every engagement ships a portable ICM packet to your storage — all of it readable by humans, replayable by any AI.

i.

Portable Packet

.zip · Markdown + manifest
Every stage, artifact, and review bundled into one zipped folder any AI can continue. Vendor-neutral. Yours.
ii.

Audit Trail

04_review/audit.md
A written human-review checkpoint at REVIEWER — the record your risk, legal, and compliance teams ask for.
iii.

Prompt Library

01_frame/_PROMPT.md
Versioned prompts and Claude Projects tuned to the employee’s recurring work — written as ICM stage briefs, diffable like code.
iv.

Agent Pipelines

03_produce/assets/
Multi-step agents wired to Slack, Notion, Drive, Salesforce — each step a named stage that can be re-run or replaced.
v.

Playbook & Roadmap

05_ship/final.md
A written SOP for how the role now operates — plus a 90-day roadmap of what to automate next.
vi.

Productivity Dashboard

packet.json telemetry
Hours-saved, quality, and adoption telemetry quotable at the board — from the packet your auditors review.
06 Tooling Philosophy Claude-first · platform-agnostic

Claude-first.
Method-portable.

We bet on Anthropic Claude as today’s best ICM executor — for reasoning quality, safety posture, agent ergonomics, and a mature ecosystem (Projects, MCP, Claude Code, Computer Use). Where another tool wins, your employee learns that too.

But the method is platform-agnostic. Packets are Markdown plus a manifest — readable by any AI you adopt. When the frontier moves, your work moves with it: no re-platforming, no re-training, no vendor holding the keys.

The result: fluency, not fanaticism. A workforce that picks the right AI — and work that outlives whichever AI wins.

Anthropic Claude
Primary platform.
Projects, sub-agents, MCP, computer use, code execution — taught at working depth.
ChatGPT & Gemini
Where they fit.
Your team learns the right model for the right job.
Cursor & Claude Code
Engineering-grade.
Engineers leave shipping at 2–4× — with reviewable, idiomatic code, not slop.
Office stack agents
Where work lives.
Slack, Notion, Drive, Excel, Salesforce — AI inside the tools they already use.
07 For Leaders Reference figures · pilot cohorts

Not another platform.
A method everyone agrees on.

Atlas is the layer between the tools your CIO selected and the workforce who must wield them: productive work, defensible quality, portable artifacts.

i.
2.6×

Productivity

Average task throughput

Knowledge-worker engagements, 30-day mark, per ICM pipeline.

ii.
9.4hrs

Quality

per employee, per week — reclaimed and auditable

Output now carries a written human-review trail.

iii.
11×

ROI

Engagement ROI in year one

Labor savings, accelerated output, avoided rework.

Pilot-cohort reference figures; measured against your own baselines.

08 Specialists 4 tracks · ICM-certified

Senior operators.
ICM-certified. Not a content team.

Every Atlas specialist has shipped production work with Claude and recertifies against the ICM curriculum on every major model release.

i.

Engineering Track

For staff engineers, platform teams, SREs.

Claude Code · Cursor · MCP · Agent SDK. ICM-structured production code, reviewable like prose.

ii.

Knowledge-Worker Track

For analysts, marketers, ops, finance, legal.

Projects · Prompt design · Document agents. Recurring role artifacts, rebuilt as ICM pipelines.

iii.

Sales & CX Track

For AEs, CSMs, support leaders.

Pipeline agents · Call summarization · CRM rituals. AI inside the deals you’re already running — every output packet-auditable.

iv.

Executive Track

For C-suite + their EAs.

Strategic AI literacy + ICM fluency. 90-minute private sessions for the policy-makers.

09 A note from the field Fortune 500 · Financial Services
We’d written off our AI rollout as a sunk cost. Atlas spent a week with one analyst — by month two her whole team was clearing 30% more deals, and our risk committee finally had something to audit. We’ve now scoped the program across four divisions.
Chief Operating Officer Fortune 500 Financial Services
10 Common Questions 8 entries

Anticipating the obvious ones.

Q1 What is the ICM Method?

ICM makes the folder structure the agent: a numbered, auditable pipeline. The implementing tool lives at icmproject.com; Atlas teaches it.

Q2 How is ICM auditable?

Every stage writes inputs, prompts, and outputs to a numbered folder; REVIEWER is a written human checkpoint.

Q3 What does portable mean — and why does it matter?

Projects export as Markdown plus a packet.json manifest, zipped — any AI continues without context loss.

Q4 How is this different from the AI training our vendor already runs?

Vendor trainings teach features; we teach a method — a packet your auditors can read and any AI can continue.

Q5 Why Anthropic Claude as the primary platform?

Claude offers the best reasoning, safety profile, and agent ecosystem today — and the method stays platform-agnostic.

Q6 How do you handle data security & IP?

Engagements run in your tenancy under your DPA; packets are your IP, written to your storage.

Q7 What’s the catch with the free hour?

None. We invest a real hour rather than pitch a deck — one employee, one pipeline, one packet.

Q8 Can we scope a department-wide pilot?

Yes — engagements graduate from one employee to a department; an ROI dashboard makes the expansion case.

11 Free Hour Zero obligation

One employee. One hour.
One packet you keep.

Pick any employee — one focused 60-minute session with a certified specialist frames their workflow as an ICM pipeline and ships the first packet.

  • ·One real recurring task, framed as an ICM pipeline
  • ·One portable packet shipped — yours to keep, runnable on any AI
  • ·Written summary & ROI projection for the requesting leader
  • ·Zero obligation. We earn the next hour.
Reserve a free hour 60 min

By submitting you agree to be contacted by an Atlas specialist. We don’t share your information.

One brief in.
One packet out.

Pick one employee, give us one hour — we ship one auditable, portable ICM pipeline and show your leadership what quality AI work looks like.

Claim a free 1-hour session → Try the method at icmproject.com ↗ Talk to a partner