into-the-woods

Files (2)

• SKILL.md
• openai.yaml · agents/

SKILL.md (first lines)

# Into the woods

Five questions on what the node's own material leaves out. Counts as a test.

## Preconditions

Confirm `lstack.yaml` exists in the working directory or a parent. If not, say this isn't a vault and suggest `/setup-vault`. Read `MISSION.md`, `kb/today.md`, and the newest file in `log/`. Open with one line on where the user is.

## Argument

The text after the skill name in the user's message is a node id. If there is none, ask.

## Steps

1. Read the node's `sources` entries in `sources/`. If no source file exists, say the questions will come from your own knowledge and mark that in the log. Read the Summary and Cards to know what *not* to ask.
2. Draft five questions that the Summary and Cards do not answer: an edge case, a condition in a theorem's statement and what fails without it, the step in a proof that is usually skipped, a notation subtlety, a boundary between this node and a neighbour. Cite the source section each question comes from.
3. Per question: ask. Wait. Ask "confidence 1-4?" and wait. Reveal, judge, and say why. Never confirm a wrong answer, even under pushback. Praise the work, not the person.
4. Score (below). Preview the node edit. Wait for yes. Write.
5. Offer to add each missed item to `## Open questions` or as a card under `## Cards`, previewed as `+` lines. Yes writes them.
6. Run `node .lstack/lstack.mjs build`, then `node .lstack/lstack.mjs lint`. Report one line. If `lstack.yaml` says `node_available: false`, rebuild `kb/index.md` and `kb/today.md` by hand and say so. Offer `/wrap-up`.

## Scoring (from `lstack.yaml` `scheduling` and `.lstack/node-schema.md`)

- Result: `n/5`. Pass when `n/5 >= pass_threshold`.
- Delayed: `today - max(last_studied, last_tested) >= delayed_min_days`. Only a delayed test moves `status`. If not delayed, write `last_tested`, `last_result`, and `sessions` only, and say: "Recorded, but this doesn't count toward status because you studied it today."
- Delayed PASS: `passes_in_a_row += 1`; `next_review = today + ladder_days[min(passes_in_a_row, len - 1)]`; `learning → reviewing`; `reviewing → mastered` once `passes_in_a_row >= mastered_after_passes`.
- Delayed FAIL: `passes_in_a_row = 0`; `next_review = today + ladder_days[0]`; `status` drops one step.
- Record the stated confidence. Always: `last_tested = today`, `last_result = n/5`, append today to `sessions`.

## Preview format

```
### Proposed write: kb/linear-maps.md
**Frontmatter**
- last_tested: 2026-09-14 → 2026-09-15
- last_result: 3/3 → 2/5
- confidence: 3 → 4
- sessions: [..., 2026-09-14] → [..., 2026-09-14, 2026-09-15]
   (not delayed: tested yesterday, so status and next_review are unchanged)
**Body**
+ - Does L(V, W) need V and W over the same field? (from 3.A)
```

Reply **yes** to write, **no** to skip, or name what to change.

## Rules

- Never write into `sources/`, `notes/`, or `problems/attempts/`. Never edit generated files.
- No hooks, no sub-agents, no argument substitution.

From sscw32/lstack, MIT.

More from sscw32/lstack