How to frame a situation

Classify a raw signal, assess capability maturity, and select the correct entry point into product shaping.

Use this when: You hold a raw signal — market observation, OKR gap, user pain, or engineering finding — and need to classify it and find the right PE-sequence entry point before committing to any solution. Prerequisites: product-engineering pack installed; a signal with a subject, an observation, and a symptom or consequence (no prior shaping artifacts required). Result: A classified signal with a Wardley capability assessment, a recommended shaping-sequence entry point, and a workspace.toml snippet to queue the item for the next step.

Goal: You have a raw signal — a market observation, an OKR gap, user pain patterns, an engineering finding — and you need to decide what kind of thing it is and where to take it next.

Skill: frame-situation (PE pack, user scope)

Frame this situation: three teams hand-roll export retention rules, so compliance reviews take weeks and produce inconsistent results.

This guide belongs to the robust shaping path.

When to use `frame-situation` vs `frame-intent`

Question	Answer	Skill
Is this affecting a whole capability area or initiative?	Yes	`frame-situation`
Is this a specific feature or screen request?	Yes	`frame-intent`
Do I know what outcome I want, just not how?	Yes	`frame-intent`
Do I have a signal but not yet know what to do with it?	Yes	`frame-situation`
Am I unsure of the altitude?	Ambiguous	Ask the skill — it will prompt you

Rule of thumb: if you can write “I want to build X so that Y” immediately, that is frame-intent territory. If you are holding an observation that something is wrong or an opportunity is emerging and you do not yet know how to respond — that is frame-situation.

What makes a well-formed signal

A well-formed signal gives the skill enough to classify. It has:

A subject — what domain or capability area the observation concerns.
An observation — something measured, seen, or reported (not a proposed solution).
A symptom or consequence — why it matters; what is happening because of it.

Good: “Three teams hand-roll their own skill discovery mechanism. No shared standard is emerging. Onboarding takes longer each quarter as a result.”

Too thin: “Skill discovery is a problem.” (No subject specificity, no consequence.) → The skill will ask for more detail.

Pre-solved: “We should build a skill registry.” → This is already a solution; frame-situation cannot classify a solution. Describe the underlying observation instead.

How to use the Wardley assessment to choose your entry point

The Wardley capability maturity table tells you how evolved each relevant capability area is. Use it to calibrate where to enter the PE six-step sequence:

What the assessment shows	Recommended entry
Capability is Genesis or early Custom-built; problem space not well-documented	Step 2 — `identify-opportunities` (map the jobs first)
Capability is mid Custom-built; problem is understood; solution options unexplored	Step 3 — `diverge-solutions`
Capability is late Custom-built or Product; options are known; need a committed bet	Step 5 — `place-bet`

When to override: The entry point is a recommendation, not a lock. If your team has already done extensive job-mapping on this problem area (it is documented somewhere), you may enter at step 3 or later. State why when you do.

When capabilities land as residual assumptions: If the agent cannot confidently place a capability on the curve, it flags it as a residual assumption. Before entering the shaping sequence, consider a quick spike (a desk-research run or team knowledge session) to ground the placement — a misread maturity can lead to over-investing in a capability already commoditized, or under-investing in one that is differentiating.

What to do with the workspace.toml suggestion

After frame-situation completes, the skill prints a TOML snippet. A dispatchable entry carries the five canonical fields — path, kind, source, summary, and needs:

{path = "docs/product/intents/agent-skill-discovery-gap.md", kind = "intent", source = {mode = "repo-origin"}, summary = "Agents cannot discover which skill fits the work in front of them", needs = []},

Add this to your active initiative’s [shaping_queue] backlog in workspace.toml. Two ways:

Ask the agent — it routes through work-intake, which writes the canonical artifact and registers the entry for you:
```
Remember the situation finding for agent skill discovery as shaping work.
```
Manual edit — open workspace.toml, find ["ini-NNN".shaping_queue], and add the entry to the backlog array.

A short {slug = "...", type = "shape"} entry is the legacy shape. It still reads during the compatibility window but is never dispatchable, so workspace-status cannot offer it as work.

Once added, workspace-status will surface it as a ready shape-typed item and suggest running identify-opportunities (or the appropriate entry-point skill).

What you have now

A classified signal with a Wardley capability assessment and a recommended entry point into the shaping sequence.
Register its canonical artifact through work-intake, then follow the recommended next shaping skill.