mirror of https://github.com/koala73/worldmonitor.git synced 2026-04-25 17:14:57 +02:00

Files

Elie Habib 110ab402c4 feat(intelligence): analytical framework selector for AI panels (#2380 )

* feat(frameworks): add settings section and import modal

- Add Analysis Frameworks group to preferences-content.ts between Intelligence and Media sections
- Per-panel active framework display (read-only, 4 panels)
- Skill library list with built-in badge, Rename and Delete actions for imported frameworks
- Import modal with two tabs: From agentskills.io (fetch + preview) and Paste JSON
- All error cases handled inline: network, domain validation, missing instructions, invalid JSON, duplicate name, instructions too long, rate limit
- Add api/skills/fetch-agentskills.ts edge function (proxy to agentskills.io)
- Add analysis-framework-store.ts (loadFrameworkLibrary, saveImportedFramework, deleteImportedFramework, renameImportedFramework, getActiveFrameworkForPanel)
- Add fw-* CSS classes to main.css matching dark panel aesthetic

* feat(panels): wire analytical framework store into InsightsPanel, CountryDeepDive, DailyMarketBrief, DeductionPanel

- InsightsPanel: append active framework to geoContext in updateFromClient(); subscribe in constructor, unsubscribe in destroy()
- CountryIntelManager: pass framework as query param to fetchCountryIntelBrief(); subscribe to re-open brief on framework change; unsubscribe in destroy()
- DataLoaderManager: add dailyBriefGeneration counter for stale-result guard; pass frameworkAppend to buildDailyMarketBrief(); subscribe to framework changes to force refresh; unsubscribe in destroy()
- daily-market-brief service: add frameworkAppend? field to BuildDailyMarketBriefOptions; append to extendedContext before summarize call
- DeductionPanel: append active framework to geoContext in handleSubmit() before RPC call

* feat(frameworks): add FrameworkSelector UI component

- Create FrameworkSelector component with premium/locked states
- Premium: select dropdown with all framework options, change triggers setActiveFrameworkForPanel
- Locked: disabled select + PRO badge, click calls showGatedCta(FREE_TIER)
- InsightsPanel: adds asterisk note (client-generated analysis hint)
- Wire into InsightsPanel, DailyMarketBriefPanel, DeductionPanel (via this.header)
- Wire into CountryDeepDivePanel header right-side (no Panel base, panel=null)
- Add framework-selector CSS to main.css

* fix(frameworks): make new proto fields optional in generated types

* fix(frameworks): extract firstMsg to satisfy strict null checks in tsconfig.api.json

* fix(docs): add blank lines around lists/headings to pass markdownlint

* fix(frameworks): add required proto string fields to call sites after make generate

* chore(review): add code review todos 041-057 for PR #2380

7 review agents (TypeScript, Security, Architecture, Performance,
Simplicity, Agent-Native, Learnings) identified 17 findings across
5 P1, 8 P2, and 4 P3 categories.

2026-03-27 23:36:44 +04:00

4.9 KiB

Raw Blame History

date, topic

date	topic
2026-03-27	intelligence-analytical-frameworks

Intelligence Analytical Frameworks

Problem Frame

WorldMonitor's AI intelligence panels (WorldBrief, Country Brief, Market Brief) generate analysis using a single neutral LLM prompt. Users who want to apply structured analytical lenses — Ray Dalio macroeconomic cycles, Buffett value frameworks, adversarial geopolitical equilibrium models — have no way to do so. Every user gets the same generic framing regardless of their decision-making context. This limits the depth and usefulness of the intelligence layer for sophisticated users and reduces differentiation for premium tiers.

Requirements

R1. Each AI intelligence panel (WorldBrief, CountryBrief, MarketBrief, and panels to be confirmed during planning) displays a framework selector in its header toolbar.
R2. The selector always includes a "Default (Neutral)" option plus all frameworks in the user's skill library.
R3. A curated built-in skill library ships with a minimum of 5 analytical frameworks, including at minimum: Ray Dalio Macroeconomic Cycles, Buffett Value & Risk Framework, and an Adversarial Geopolitical Equilibrium model. The full list is defined during planning.
R4. When an active framework is selected, its analytical instructions augment the LLM prompt for that panel's intelligence generation — shaping how events are interpreted through that lens.
R5. Switching a framework on a panel immediately triggers a fresh analysis with the new lens applied; the panel updates in place.
R6. Framework selection per panel persists across sessions.
R7. Premium users can import additional frameworks from agentskills.io by entering a skill URL/ID or by pasting a compatible skill definition; imported skills are saved to the user's skill library.
R8. The skill library is accessible and manageable (view, rename, delete) from the Settings panel.
R9. Non-premium users see the default neutral analysis only. The framework selector is visible but locked, with an upgrade prompt.

Success Criteria

A premium user can switch CountryBriefPanel to the Ray Dalio framework and receive an analysis structured around economic seasons, risk parity, and root-cause diagnosis rather than neutral summarization.
A custom skill can be imported from agentskills.io, saved to the library, and applied to a panel within the same session.
Switching frameworks triggers a re-analysis within the same latency window as a normal panel refresh.
Non-premium users encounter a locked selector with a clear upgrade CTA — not a hidden feature.

Scope Boundaries

agentskills.io tools (callable functions) are not executed in this phase; only the instructions field is used for prompt augmentation.
Framework selection applies only to AI intelligence panels, not to MCP data panels.
No server-side skill execution in this phase.
No community marketplace or cross-user skill sharing in this phase.
One framework active per panel at a time (no stacking).

Key Decisions

Per-panel selection over global: Different panels serve different contexts (macro vs. country vs. market); a single global lens would be too blunt.
Auto re-run on switch: Waiting for a manual refresh creates confusion about whether the framework is applied; immediate re-analysis makes the effect visible and the feature feel alive.
Premium only: Framework extensibility is a differentiated capability that justifies the premium tier; neutral analysis remains available to all.
Instructions-only from agentskills.io (phase 1): Tool/function execution requires server-side orchestration and is a meaningful scope expansion; deferring keeps this phase lightweight and shippable.

Dependencies / Assumptions

The existing summarization/LLM prompt chain supports prompt augmentation at the panel level (to be confirmed during planning).
agentskills.io skills can be fetched or pasted as JSON/YAML and their instructions field extracted.

Outstanding Questions

Resolve Before Planning

(none blocking — proceed to planning)

Deferred to Planning

[Affects R1][Technical] Which additional panels beyond WorldBrief/CountryBrief/MarketBrief should support framework selection? (DeductionPanel, GdeltIntelPanel candidates)
[Affects R4][Technical] Where in the LLM prompt chain is the framework injected? (system prompt prepend, separate layer, or alongside existing instructions)
[Affects R7][Needs research] How are imported skills persisted for premium users — localStorage only, or synced to backend? Evaluate cost/complexity tradeoff.
[Affects R3][Needs research] Finalize the built-in framework list: confirm Ray Dalio, Buffett, Geopolitical Equilibrium; evaluate Sun Tzu / conflict theory, psychohistory (Seldon/Dalio hybrid), and others from issue #2291.

Next Steps

→ /ce:plan for structured implementation planning

4.9 KiB Raw Blame History