Six upright ring arches in a row, each holding one fine probe over a chute dropping a cut chip, beside a rail carrying a faceted specimen block, cut diagonally so the far three arches are drawn as bone-grey outlines, with a blue collector channel running past every chute into a ring terminal.

Other

IterativeDepth

Structured multi-angle exploration running 2-8 sequential passes over the same problem, each through a different scientific lens, to surface hidden requirements and edge cases invisible from one angle; each pass yields new ISC criteria

Workflow

Tools

References

Triggers

The Problem

When you ask an AI to analyze a problem, it runs one pass and gives you the obvious answer — the surface requirements, the happy path, the constraints the user mentioned. That's it. Everything invisible from that single angle stays invisible: the stakeholder who wasn't in the room, the failure mode that only shows up under load, the implicit assumption that makes the whole approach wrong in six months. Single-pass analysis is fast, but the missed requirements are what blow up mid-execution.

How This Skill Approaches It

IterativeDepth runs 2–8 sequential passes through the same problem, each from a structurally different lens. Lens 1 extracts what was literally said. Lens 2 asks who else is affected. Lens 3 assumes the solution shipped and looks for what breaks. Lens 4 examines what changes over time. Lens 5 asks how it should feel. Lens 6 inverts constraints to surface hidden assumptions. Each pass outputs new ISC criteria the previous passes couldn't see — the Explore workflow runs all passes, stops when yields repeat, and feeds the combined criteria into the ISA. The technique draws from validated work across 20 frameworks: Gadamer's hermeneutic circle for progressive pre-understanding, de Bono's Six Thinking Hats for structured angle rotation, Wang et al.'s Self-Consistency for why multiple reasoning paths beat one, and Finkelstein's Viewpoint-Oriented Requirements Engineering for the stakeholder lens.

Not for scope/zoom analysis (use ApertureOscillation)

In Action

What you say to your DA, and what the IterativeDepth skill actually does.

You say "use iterative depth on this api redesign before we write specs"

Runs the Explore workflow with up to 6 lenses on the API problem — extracting stated requirements, stakeholder needs, failure modes, temporal concerns, experiential quality, and hidden assumptions — then outputs a unified set of ISC criteria covering what a single-pass analysis would miss.
You say "quick depth on this, just two angles"

Runs Explore in Fast mode (2 lenses), applying the Literal and Failure lenses to surface the most common missing criteria without the full time budget of a complete multi-pass run.

Inside the Skill

The thinking, frameworks, and architecture that distinguish this skill from a generic version of the same task.

Ideal State

A run returns ISC criteria and anti-criteria that a single pass over the same problem would have missed, each one binary-testable and deduplicated across lenses.
The lenses fit the problem (a security problem leans on the failure and adversary angles, a UX problem on the experiential one). A run uses between two and eight passes and ends once a pass only repeats earlier findings, usually by about the fifth.
The result names the lenses used and the cross-angle finding that only appeared because two lenses collided.

Intent Routing

Intent	Trigger	File
Explore	"iterative depth", "explore deeper", "multi-angle"	`Intents/Explore.md`

File	Read when
`TheLenses.md`	choosing lenses: the catalog of eight exploration angles, concrete to abstract, each with its question, grounding and example prompt
`ScientificFoundation.md`	justifying the method or explaining why multi-lens passes work: the 20 techniques it rests on and how it differs from CS iterative deepening

Data & Config

User customizations live at <configRoot>/LIFEOS/USER/CUSTOMIZATIONS/SKILLS/IterativeDepth/ (configRoot is $CLAUDE_CONFIG_DIR when set, else ~/.claude). When that directory exists, its PREFERENCES.md, configurations and resources override the defaults; when it is absent, the defaults stand.

Examples

User: "Inline, no agents: 3 quick passes over the issue-security route design (private vs public)"
→ Explore intent
→ Three lenses run inline; new criteria and anti-criteria for each route outcome, with the cross-angle finding

User: "What am I missing in this ISA coordination design?"
→ Explore intent
→ Criteria and anti-criteria the design lacked and existing criteria refined, with the lenses named, returned to the user

Workflows & Routing · 1

Each workflow is one job the skill runs. The trigger phrases route your request to the right one — this is the skill's routing table.

01

Explore Intents/Explore.md

iterative depth, explore deeper, multi-angle

How to Invoke

Say any of these to your DA and LifeOS activates the IterativeDepth skill automatically:

"iterative depth"
"explore deeper"
"multi-angle analysis"
"surface hidden requirements"
"blind spot check"
"what am I missing"

Or invoke explicitly:

Skill("IterativeDepth")

References · 2

Auxiliary files the skill loads at runtime — frameworks, guides, configs.

ScientificFoundation
TheLenses

Related Skills

References & Credits

The thinkers, books, frameworks, and research this skill is built on. The ideas belong to them — the integration belongs to LifeOS.

Authors

Framework

Causal Layered Analysis Sohail Inayatullah's four-layer futures framework — one of the deep-dive lenses.

Research

Self-Consistency Prompting Wang et al. (2022) showed multiple diverse reasoning paths beat single-path chain-of-thought; IterativeDepth applies the same principle to requirements extraction.

Want LifeOS to do this for you?

Install LifeOS on your machine — your DA gets the IterativeDepth skill plus 96 others, all hooked into one Life OS.

Install LifeOS View on GitHub