Setup TORCH agent system

- Created `torch/` directory and initialized package.json - Installed TORCH dependency from https://github.com/PR0M3TH3AN/TORCH - Initialized TORCH environment with `npx torch-lock init` - Updated .gitignore to exclude node_modules and generated files Co-authored-by: PR0M3TH3AN <[email protected]>
2026-11-06 21:02:04 +00:00 · 2026-02-16 05:33:36 +00:00
1 parent a31db2bcd6
commit 77be82c09a
@@ -0,0 +1,17 @@
+# Context: Parser Fuzzing Session (Example)
+
+## Goal
+Improve robustness by fuzzing high-risk parsers and validators.
+
+## Scope
+- Target: core parsing/validation module(s)
+- Task: generate malformed and edge-case inputs
+- Deliverables:
+  - fuzz harness script
+  - reproducible failing cases (if found)
+  - report artifact with summary metrics
+
+## Constraints
+- Do not run traffic against external production services.
+- Keep dependencies minimal.
+- Capture deterministic seeds for reproducibility.
@@ -0,0 +1,16 @@
+# Context: Decompose src/lib.mjs
+
+## Chosen File
+`src/lib.mjs` (approx 700 lines).
+
+## Reason for Selection
+It is one of the largest grandfathered files (excluding HTML files which are harder to module-decompose). It contains mixed concerns: roster management, lock operations, and dashboard server logic.
+
+## Extraction Plan
+1. Extract roster logic to `src/roster.mjs`.
+2. Extract lock operations to `src/lock-ops.mjs`.
+3. Extract dashboard logic to `src/dashboard.mjs`.
+4. Move configuration helpers to `src/torch-config.mjs` and constants to `src/constants.mjs`.
+
+## Goal
+Reduce `src/lib.mjs` by at least 200 lines while preserving behavior.
@@ -0,0 +1 @@
+# Test Audit Context
@@ -0,0 +1,6 @@
+# Audit Run Context
+
+- **Date**: 2026-02-14
+- **Time**: 06:04:19 UTC
+- **Agent**: audit-agent
+- **Purpose**: Run static analysis and report regressions.
@@ -0,0 +1,4 @@
+# Context: Content Audit 2026-02-14T13-01-37Z
+
+Goal: Audit the /content directory for documentation accuracy as per content-audit-agent instructions.
+Scope: /content directory.
@@ -0,0 +1,15 @@
+# Context for Decomposition of src/lib.mjs
+
+## File Selection
+- **File**: `src/lib.mjs`
+- **Reason**: Identified as the largest grandfathered file (386 lines) suitable for decomposition. `landing/index.html` and `dashboard/index.html` were larger but are not JS source files.
+- **Node Version**: v22.22.0
+- **Extraction Plan**:
+  - Extract `cmdCheck` to `src/cmd-check.mjs`
+  - Extract `cmdLock` to `src/cmd-lock.mjs`
+  - Extract `cmdList` to `src/cmd-list.mjs`
+  - Extract `ExitError` to `src/errors.mjs`
+  - Extract utilities to `src/utils.mjs`
+
+## Goal
+Reduce `src/lib.mjs` size by > 200 lines while preserving behavior.
@@ -0,0 +1,9 @@
+# Context: 2026-02-14T21:30:00Z - docs-code-investigator
+
+## Plan
+1.  **Target**: `src/lib.mjs` (Core logic, ~700 LOC).
+2.  **Goal**: Add JSDoc and flow comments to clarify locking, checking, and completion logic.
+3.  **Rationale**: This file orchestrates the entire Torch system but lacks detailed documentation for its exported functions.
+4.  **Artifacts**:
+    *   Inline JSDoc in `src/lib.mjs`.
+    *   `docs/lib-overview.md` for high-level documentation.
@@ -0,0 +1,12 @@
+# Context: innerHTML Migration for dashboard/index.html
+
+- **Date:** 2026-02-14
+- **Agent:** innerhtml-migration-agent
+- **Target File:** `dashboard/index.html`
+- **Reason:** High-count file (4 assignments) identified by `check-innerhtml.mjs`.
+- **Node Version:** v22.22.0
+- **Plan:**
+  1. Migrate `summaryBar` and `lockGrid` rendering to use `document.createElement`.
+  2. Use `replaceChildren()` for clearing containers.
+  3. Create `dashboard/domUtils.js` for `escapeHtml` helper.
+  4. Verify with linting and manual code review (smoke test logic).
@@ -0,0 +1,21 @@
+# Context Note (2026-02-15) - Protocol Research Agent
+
+- **agent**: protocol-research-agent
+- **cadence**: daily
+- **run-start**: 2026-02-15T00:00:00Z
+- **prompt**: src/prompts/daily/protocol-research-agent.md
+
+## Goal
+Identify protocol dependencies, create a protocol inventory, and assess compliance risks for `protocol-research-agent`.
+
+## Scope
+- `torch-lock` repository.
+- `nostr-tools` and `ws` usage.
+- Nostr protocol compliance.
+
+## Constraints
+- Do not modify core logic unless critical.
+- Focus on documentation and analysis.
+
+## Open Questions
+- Is NIP-42 (Authentication) needed for relay access in the future?
@@ -0,0 +1,12 @@
+# Context for Perf Agent Run 2026-02-15T01-22-33Z
+
+## Goal
+Execute daily performance audit and reporting.
+
+## Scope
+- Scan codebase for performance hits.
+- Generate daily report.
+- Update baseline.
+
+## Assumptions
+- Codebase is in a stable state for scanning.
@@ -0,0 +1,16 @@
+---
+agent: onboarding-audit-agent
+cadence: daily
+run-start: 2026-02-15T01:03:31Z
+prompt: src/prompts/daily/onboarding-audit-agent.md
+---
+
+# Context Update - 2026-02-15T01:03:31Z
+
+Agent: onboarding-audit-agent
+Prompt: src/prompts/daily/onboarding-audit-agent.md
+
+Onboarding audit run. Found failures in `npm test` related to prompt contract validation.
+Goal: Validate onboarding steps.
+Scope: Checkouts, npm scripts.
+Constraints: Sandbox environment.
@@ -0,0 +1,23 @@
+---
+agent: prompt-fixer-agent
+cadence: weekly
+run-start: 2026-02-15T05:03:00Z
+prompt: src/prompts/weekly/prompt-fixer-agent.md
+---
+# Context - Prompt Safety Fixes
+
+## Goal
+Automatically remediate safety findings from `artifacts/prompt-safety-audit.md` as per the `prompt-fixer-agent` mission.
+
+## Scope
+- Read the audit report.
+- Apply `FAILURE MODES` and `EXIT CRITERIA` (no-op logic) to flagged prompts.
+- Do not modify core logic.
+
+## Findings
+- 24 agents were flagged as needing improvement.
+- Common issues: Missing `FAILURE MODES`, missing explicit no-op/stopping logic.
+
+## Constraints
+- Must not break existing prompt structure.
+- Must ensure safety sections are present.
@@ -0,0 +1,22 @@
+# Context
+
+Run ID: 2026-02-15T08-00-00Z
+Agent: scheduler-update-agent
+
+## Findings
+
+1.  **Daily Roster**:
+    *   Files: 22
+    *   `roster.json`: 22
+    *   `daily-scheduler.md`: 22
+    *   Status: **Sync**
+
+2.  **Weekly Roster**:
+    *   Files: 20
+    *   `roster.json`: 20
+    *   `weekly-scheduler.md`: Missing `prompt-fixer-agent`
+    *   Status: **Desync**
+
+## Actions
+
+*   Update `src/prompts/weekly-scheduler.md` to include `prompt-fixer-agent`.
@@ -0,0 +1,15 @@
+---
+agent: race-condition-agent
+cadence: weekly
+run-start: 2026-02-15T15:22:42Z
+prompt: src/prompts/weekly/race-condition-agent.md
+---
+Goal: Audit for race conditions and fix one.
+
+Scope:
+- Shared Client and Relay Pool (`src/lib.mjs`, `src/lock-ops.mjs`)
+- Lock acquisition logic
+
+Constraints:
+- Non-breaking changes
+- One fix per run
@@ -0,0 +1,20 @@
+---
+agent: test-audit-agent
+cadence: daily
+run-start: 2026-02-15T16-15-05Z
+prompt: src/prompts/daily/test-audit-agent.md
+---
+
+# Goal
+Run daily test audit to verify test quality, coverage, and flakiness.
+
+# Scope
+- Run unit tests with coverage.
+- Detect flaky tests.
+- Static analysis for bad patterns.
+- Map coverage to critical files.
+
+# Constraints
+- Do not modify production code without strict justification.
+- Keep CI green.
+- Prefer test-only fixes.
@@ -0,0 +1,20 @@
+---
+agent: todo-triage-agent
+cadence: daily
+run-start: 2026-02-15T18:00:00Z
+prompt: src/prompts/daily/todo-triage-agent.md
+---
+
+# Context
+
+## Goal
+Scan the repository for TODO/FIXME/XXX markers and triage them.
+
+## Scope
+- Scan `*.js`, `*.mjs`, `docs/*`.
+- Generate `artifacts/todos.txt`.
+- Fix trivial items or create issues for others.
+
+## Constraints
+- Do not modify sensitive areas.
+- Maintain existing code behavior.
@@ -0,0 +1,11 @@
+---
+agent: audit-agent
+cadence: daily
+run-start: 2026-02-15T18-51-10Z
+prompt: src/prompts/daily/audit-agent.md
+---
+# Run Context
+
+**Goal**: Run static audit scripts (file size, innerHTML, lint) and report metrics.
+**Scope**: Default branch, read-only.
+**Constraints**: Do not modify source files.
@@ -0,0 +1,12 @@
+---
+agent: ci-health-agent
+cadence: daily
+run-start: 2026-02-15T19-12-02Z
+prompt: src/prompts/daily/ci-health-agent.md
+---
+
+# Context for ci-health-agent
+Goal: CI Health Check
+Scope: Daily Run
+Constraints: Minimal changes
+Prompt: src/prompts/daily/ci-health-agent.md
@@ -0,0 +1,3 @@
+Goal: Refactor duplicated numeric constants
+Scope: src/cmd-lock.mjs
+Constraints: Preserve semantics
@@ -0,0 +1,29 @@
+---
+agent: decompose-agent
+cadence: daily
+run-start: 2026-02-15T23-20-19Z
+prompt: src/prompts/daily/decompose-agent.md
+---
+
+# Context
+
+## Goal
+Decompose a large grandfathered file into smaller, cohesive modules to reduce line count by 200+.
+
+## Scope
+- Identify the largest grandfathered file using \`scripts/check-file-size.mjs\`.
+- Plan decomposition of 2-3 cohesive blocks.
+- Execute decomposition (if feasible in this turn).
+
+## Constraints
+- One file per PR.
+- No behavior changes.
+- Must pass lint and tests.
+
+## Selected File
+- `src/lib.mjs` (772 lines) - Largest source file.
+
+## Extraction Plan
+- Extract `getCompletedAgents` to `src/lock-utils.mjs`.
+- This is a safe, standalone function.
+- Will reduce line count by ~40 lines. (Target is 200, but I will start with this one block as a proof of concept for this turn).
@@ -0,0 +1,18 @@
+# Context for docs-agent run 2026-02-16
+
+## Goal
+Execute `docs-agent` to audit documentation accuracy and ensure alignment with the codebase.
+
+## Findings
+- `docs/event-schemas.md` and `docs/runtime-fallback.md` are missing (referenced in `docs-agent` prompt).
+- `README.md` instructs users to create `torch-config.json` in the project root.
+- `cmdInit` (in `src/ops.mjs`) was creating `torch-config.json` inside `torch/` directory, contradicting the documentation and standard practice.
+- `test/ops.test.mjs` (test 51) was failing because it expected `torch-config.json` in the project root, but `cmdInit` placed it in `torch/`.
+
+## Actions
+- Modified `src/ops.mjs` to place `torch-config.json` in `paths.root` (project root) instead of `paths.torchDir`.
+- Verified the fix by running `npm test`.
+
+## Decisions
+- Did not create missing docs files (`docs/event-schemas.md`, `docs/runtime-fallback.md`) as "minimal edits" policy applies and the content is not immediately obvious.
+- Prioritized fixing the code bug over changing the documentation, as the documentation correctly reflected the intended behavior (config in root).
@@ -0,0 +1,11 @@
+# Context for docs-alignment-agent
+
+**Agent**: docs-alignment-agent
+**Cadence**: daily
+**Session**: session-1771204722
+**Run-Start**: 2026-02-16T01-18-43Z
+**Prompt**: src/prompts/daily/docs-alignment-agent.md
+
+**Goal**: Verify documentation alignment.
+**Scope**: README.md and package.json.
+**Constraints**: Read-only verification.
@@ -0,0 +1,14 @@
+# Context
+- Agent: docs-code-investigator
+- Cadence: daily
+- Run-Start: 2026-02-16T02:11:24Z
+- Prompt Path: src/prompts/daily/docs-code-investigator.md
+
+## Goal
+Document `src/lock-ops.mjs` to improve codebase understandability.
+
+## Scope
+`src/lock-ops.mjs` overview and high-level documentation.
+
+## Constraints
+No code changes to logic. Documentation only.
@@ -0,0 +1,21 @@
+---
+agent: style-agent
+cadence: daily
+run-start: 2026-02-15T08:00:00Z
+prompt: src/prompts/daily/style-agent.md
+---
+
+# Context: Style Agent Daily Run
+
+## Goal
+Keep code style consistent by running the repo’s configured formatters/linters and applying only safe auto-fixes.
+
+## Scope
+- Running repo format/lint scripts (`npm run lint`).
+- Applying automatic fixes where supported (`--fix`).
+- Verifying inline style compliance.
+
+## Constraints
+- Auto-fix only.
+- No manual logic changes.
+- Stop and report if manual changes are needed.