feat(self-modification): backend engine, REST API, and frontend panel

[Copilot]

The backend self-modification engine referenced by the existing frontend stores was absent. This PR delivers the engine, wires it to four new REST endpoints, and adds a UI panel that drives the full propose → apply → rollback lifecycle.

Backend: SelfModificationEngine

New backend/core/self_modification_engine.py — standalone, no heavy dependencies:

engine = SelfModificationEngine(cognitive_manager)   # cognitive_manager optional

result = engine.propose_modification(
    target="knowledge_graph",
    modification_type="add_node",
    parameters={"node_id": "foo", "node_data": {"label": "Foo"}},
)
# → ProposalResult(proposal_id=uuid, risk_level="low", predicted_impact={...}, status="pending")

engine.apply_modification(result.proposal_id)     # mutates live NetworkX graph
engine.rollback_modification(result.proposal_id)  # snapshot-based revert
engine.get_modification_history()                 # full audit trail, newest-first

Supported targets: knowledge_graph (add/remove nodes + edges), reasoning_params (parameter tuning), goal_priority (goal reweighting). Operates in simulation mode when no CognitiveManager is provided.

REST API (unified_server.py)

Method	Path	Returns
POST	/api/self-modification/propose	proposal_id, predicted_impact, risk_level
POST	/api/self-modification/apply/{proposal_id}	success, applied_changes
POST	/api/self-modification/rollback/{proposal_id}	success, reverted_to
GET	/api/self-modification/history	history[], total

Guarded by SELF_MODIFICATION_AVAILABLE — returns 503 cleanly if the module fails to import.

Frontend

• SelfModificationPanel.svelte — 3-tab panel: Pending Proposals (risk badges, apply action), History (full audit table with rollback action), New Proposal (target/type dropdowns with JSON parameter editor and auto-populated templates).
• App.svelte — panel registered as 🔧 Self-Modification in the System Management nav section (featured: true).
• Syncs pending proposals to the existing evolutionState store (pendingProposals derived store).

Tests

tests/backend/test_self_modification_engine.py — 27 tests covering the four required scenarios:

• test_propose_modification_returns_proposal_id
• test_apply_modification_changes_knowledge_graph
• test_rollback_reverts_applied_modification
• test_history_records_all_operations

Related Issues

Test Evidence

tests/backend/test_self_modification_engine.py  27 passed in 0.47s

Frontend: npm run build — ✓ built in 6.31s, 0 errors.

Checklist

• Tests pass locally (pytest tests/)
• Code is formatted (black . and isort .)
• Documentation updated (if applicable)
• No secrets or credentials committed
• Related issue linked above

Original prompt

Context — Issue #95

PR #41 implements a self-modification UI panel in the Svelte frontend, but the backend engine it is intended to drive is either dormant or absent. This issue delivers the backend self-modification engine and wires it to the existing UI.

Tasks

1. Audit existing self-modification primitives

• Search godelOS/ for any existing self-modification, meta-learning, or architecture-modification classes/functions
• Identify what can be reused vs what must be implemented fresh

2. Implement SelfModificationEngine

Create backend/core/self_modification_engine.py (or activate the dormant equivalent) with:

• propose_modification(target: str, modification_type: str, parameters: dict) -> ProposalResult — creates a sandboxed proposal with predicted impact
• apply_modification(proposal_id: str) -> ApplyResult — applies the accepted proposal to the live knowledge graph or reasoning module configuration
• rollback_modification(proposal_id: str) -> RollbackResult — reverts an applied modification
• get_modification_history() -> list[ModificationRecord] — returns audit trail of all modifications

Modification targets must include at minimum:

• Knowledge graph assertion (add/remove nodes or edges)
• Reasoning module parameter tuning (e.g., inference depth, timeout)
• Active goal priority reweighting

3. REST API endpoints

Add to backend/unified_server.py (or appropriate router file):

• POST /api/self-modification/propose — body: { target, modification_type, parameters } → returns { proposal_id, predicted_impact, risk_level }
• POST /api/self-modification/apply/{proposal_id} → returns { success, applied_changes }
• POST /api/self-modification/rollback/{proposal_id} → returns { success, reverted_to }
• GET /api/self-modification/history → returns list of ModificationRecord

4. Connect to the frontend (PR #41 UI)

• Ensure the API response shapes match what the existing Svelte self-modification panel expects (inspect the existing frontend component to determine the contract)

5. Tests

• test_propose_modification_returns_proposal_id
• test_apply_modification_changes_knowledge_graph
• test_rollback_reverts_applied_modification
• test_history_records_all_operations

Acceptance Criteria

Self-modification operations initiated via the UI produce verifiable changes to the knowledge graph or active reasoning modules, and can be rolled back.

Related Issue

Fixes #95

This pull request was created from Copilot chat.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.