FE-698: Prompt/context substrate planning

[lunelson]

Stack Context

This FE-698 stack establishes the prompt/context scenario substrate that future observer, interviewer, research, behavioral-kernel, architect, and decomposition probes can share before product UI is committed.

What?

• Updates memory/SPEC.md with prompt/context substrate requirements, assumptions, and decisions.
• Updates memory/PLAN.md to make FE-698 the next prompt/context frontier with downstream traceability.
• Adds the initial scoped card for packaged markdown prompt assets.

Why?

Multi-chat and graph-first agent features need shared prompt/context machinery instead of each call site inventing its own transcript/context hack. This planning PR creates the canonical baseline for the implementation PRs stacked above it.

[lunelson]

• FE-705: Plan agent capability CLI frontier #127
• FE-698: On more context pack + spec for agent CLI v1 #114
• FE-698: Prompt/context scenario substrate foundation #111
• FE-698: Agent mutation audit; tool surface design #104
• FE-698: Design/spec/plan augmentation #103
• FE-698: Scenario runner + observer scenario composition #102
• FE-698: Prompt registry + observer context-pack foundation #101
• FE-698: Prompt/context substrate planning #98 👈 (View in Graphite)
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[cursor]

PR Summary

Low Risk
Documentation-only changes that restructure planning/spec text; no runtime code paths or data handling are modified.

Overview
Defines the FE-698 prompt/context scenario substrate as the next frontier by updating memory/PLAN.md and memory/SPEC.md with new requirements/decisions/assumptions around packaged markdown prompts, typed context packs, a scenario runner for pre-UI prompt probes, and a Pi harness spike boundary.

Adds memory/CARDS.md to scope the first implementation slice (packaged markdown prompt registry), including acceptance criteria and test/packaging risks focused on preserving existing prompt text while moving it out of TypeScript.

^{Reviewed by Cursor Bugbot for commit e4237c5. Bugbot is set up for automated code reviews on this repo. Configure here.}

[augmentcode]

🤖 Augment PR Summary

Summary: Planning-only PR for FE-698 to establish a shared baseline for the upcoming prompt/context “scenario substrate” work (before implementation branches stack on top).

Changes:

• Added memory/CARDS.md with an initial scope card for a packaged markdown prompt registry (typed loader/registry + prompt-text preservation tests + packaging boundary verification).
• Reordered and expanded memory/PLAN.md to pull “Prompt/context scenario substrate” into Next (Implement sql version-control with Dolt DB #2) and clarify downstream dependencies (ontology/progressive-checkability, prompt probes before UI, web research, candidate-spec assist, continuous workspace).
• Updated memory/SPEC.md to make prompt/context engineering explicit in the product direction, and introduced new traceability anchors: Requirements 40–41, Assumptions A84–A87, Decisions D139–D142, and Invariant I112.

Technical Notes: The plan emphasizes scenario-specific graph context packs over transcript-dump context, plus a lightweight scenario runner to capture rendered prompts/context, model settings, raw output, and structured parse status; Pi is positioned as a potential harness adapter rather than product runtime authority.

_{🤖 Was this summary useful? React with 👍 or 👎}

[augmentcode]

Review completed. No suggestions at this time.

Comment augment review to trigger a new review at any time.