Hybrid Research Rig v0.1.3 — ensemble bot, Baritone pathfinding, vision, HUD, security hardening by Z0mb13V1 · Pull Request #714 · mindcraft-bots/mindcraft

Z0mb13V1 · 2026-03-01T13:18:50Z

3/3/2026
PR #716 consolidated everything from #710, #714, #717, and #718.
This PR was superseded by #716.

Copilot

Pull request overview

This PR brings in the “Hybrid Research Rig v0.1.3” feature set: multi-model ensemble decision-making, Baritone-based pathfinding, vision/HUD support, remote MindServer mode, and broad security + resilience hardening across agent lifecycle, messaging, and persistence.

Changes:

Add an ensemble model pipeline (panel + heuristic arbiter + optional LLM judge) with logging and ChromaDB-backed experience retrieval.
Replace/extend navigation and vision subsystems (Baritone integration, safer async camera/viewer loading) and strengthen agent process resilience + persistence (atomic writes, restart backoff).
Security hardening across input validation, rate limiting, key loading precedence, sandbox notes, and safer TTS execution.

Reviewed changes

Copilot reviewed 133 out of 134 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
tasks/human_evaluation.js	Minor lint/unused-var adjustments for task runner script.
tasks/construction_tasks/get_blueprint.js	Minor lint/format fixes in blueprint generation helper.
tasks/dragon/diamond_pickaxe.json	New dragon progression task chunk config.
tasks/dragon/nether_portal.json	New dragon progression task chunk config.
tasks/dragon/blaze_rods.json	New dragon progression task chunk config.
tasks/dragon/ender_pearls.json	New dragon progression task chunk config.
tasks/dragon/stronghold.json	New dragon progression task chunk config.
tasks/dragon/ender_dragon.json	New dragon progression task chunk config.
tasks/dragon/full_run.json	New “full run” dragon progression task config.
start.ps1	Add Windows launcher for starting both bots via Docker Compose profiles.
src/utils/usage_tracker.js	New usage/cost tracking with periodic autosave + RPM/TPM snapshot.
src/utils/rate_limiter.js	New in-memory rate limiter with cleanup interval + destroy hook.
src/utils/message_validator.js	New Discord/Minecraft message + username validation/sanitization helpers.
src/utils/mcdata.js	Switch auto-eat loader style; load Baritone; configure ashfinder defaults on spawn.
src/utils/keys.js	Prefer env vars over keys.json; sanitize loaded keys; warn on legacy usage.
src/utils/examples.js	Minor lint/format adjustments.
src/process/init_agent.js	Add remote MindServer connection mode; add global exception/rejection handlers.
src/process/agent_process.js	Support remote agent mode; improve restart/backoff behavior; use `process.execPath`.
src/models/vllm.js	Refactor vLLM client; retries; usage capture; basic `<think>` stripping; add vision wrapper.
src/models/qwen.js	Remove unused `hasKey` import.
src/models/openrouter.js	Remove unused `hasKey` import; embed stub arg rename.
src/models/ollama.js	Add usage capture; replace `fetch` usage with http/https request; Ollama-native vision formatting.
src/models/novita.js	Embed stub arg rename.
src/models/mercury.js	Remove unused `hasKey` import.
src/models/huggingface.js	Embed stub arg rename.
src/models/grok.js	Add usage capture; null-safe separator replacement; embed stub arg rename.
src/models/gpt.js	Minor lint-variable renaming inside OpenAI provider.
src/models/glhf.js	Normalize line endings/formatting; embed stub arg rename.
src/models/gemini.js	Capture usage metadata; adjust generateContent config shape; embed return compatibility.
src/models/deepseek.js	Remove unused `hasKey` import; embed stub arg rename.
src/models/claude.js	Embed stub arg rename.
src/models/cerebras.js	Stop-seq arg rename; embed stub arg rename.
src/models/_model_map.js	Guard against null/undefined model profile selection.
src/mindcraft/mindcraft.js	Add remote agent creation (writes settings file, spawns remote-connected agent process).
src/ensemble/panel.js	New panel fan-out with per-model timeout and proposal parsing.
src/ensemble/arbiter.js	New heuristic scoring + majority bonus + latency penalty + confidence calculation.
src/ensemble/judge.js	New LLM-as-judge fallback for low-confidence heuristic outcomes.
src/ensemble/logger.js	New decision log writer with trimming + basic stats.
src/ensemble/feedback.js	New ChromaDB feedback collector for “similar situations” memory injection + outcome updates.
src/ensemble/controller.js	New EnsembleModel drop-in implementing phased decision pipeline + usage aggregation.
src/agent/vision/vision_interpreter.js	Switch to async camera import with graceful fallback on missing native deps/WebGL.
src/agent/vision/camera.js	Fix viewer init ordering; add readiness flag + destroy method; improve error handling.
src/agent/vision/browser_viewer.js	Lazy-load prismarine-viewer; bind host; degrade gracefully when module missing.
src/agent/tasks/tasks.js	Minor linting/format; small control-flow cleanup.
src/agent/tasks/cooking_tasks.js	Minor linting/format; rename unused locals.
src/agent/speak.js	Replace shell-based TTS with `spawn` arg arrays/stdin to mitigate injection; async queue handling.
src/agent/self_prompter.js	Minor formatting/semicolons/log statements.
src/agent/npc/item_goal.js	Minor lint/semicolons; ignore unused tuple element.
src/agent/npc/controller.js	Harden construction JSON loading (resolved dir + path traversal guard); minor lint.
src/agent/npc/build_goal.js	Remove unused mcdata import.
src/agent/mindserver_proxy.js	Add remote registration flow; connection timeout; ignore unnamed/broadcast restart events; add usage RPC.
src/agent/memory_bank.js	Minor semicolon fixes.
src/agent/library/world.js	Switch clear-path checks to Baritone generatePath; remove pathfinder dependency.
src/agent/library/skill_library.js	Improve embedding-model error logging; minor lint.
src/agent/library/lockdown.js	Clarify unsafeEval rationale; minor export formatting.
src/agent/library/index.js	Add dragon_runner docs to skill doc generation.
src/agent/library/full_state.js	Include resumeName + selfPrompter state in full state snapshot.
src/agent/learnings.js	New persistent “learnings” store with atomic write + stats/summary helpers.
src/agent/history.js	Atomic writes; sanitize stale “broken gathering” beliefs; guard corrupted/empty files; persist learnings.
src/agent/conversation.js	Make endConversation/endAllConversations async; await scheduling; minor semicolons.
src/agent/connection_handler.js	Allow empty catch blocks; minor lint.
src/agent/commands/queries.js	Minor formatting; string escape fix; add semicolons.
src/agent/commands/index.js	Track blocked commands; regex semicolon; minor error message formatting.
src/agent/coder.js	Use sync template reads; add optional code execution timeout; fix missing-skills check; minor logging fixes.
src/agent/action_manager.js	Add stuck/loop detection and collect-fail heuristics; adjust stop behavior; improve error-stack logging.
settings.js	Add deepSanitize for SETTINGS_JSON; update defaults for hybrid rig use (vision, prompts, profiles, etc.).
services/viaproxy/README.md	Formatting and list cleanup.
profiles/vllm.json	Add mode scaffolding + local URL changes.
profiles/qwen.json	Add mode scaffolding + hybrid/local variants.
profiles/mistral.json	Add mode scaffolding + hybrid/local variants.
profiles/mercury.json	Add mode scaffolding + hybrid/local variants.
profiles/llama.json	Add mode scaffolding + hybrid/local variants.
profiles/grok.json	Update identity/prompting; add mode scaffolding.
profiles/gpt.json	Update default OpenAI model; add mode scaffolding.
profiles/gemini.json	Update identity/prompting; add mode scaffolding; embedding selection.
profiles/deepseek.json	Add mode scaffolding + hybrid/local variants.
profiles/claude_thinker.json	Add mode scaffolding + hybrid/local variants.
profiles/claude.json	Add mode scaffolding + prompt changes.
profiles/andy-4.json	Add mode scaffolding + hybrid/local variants.
profiles/andy-4-reasoning.json	Add mode scaffolding + prompt tweaks.
profiles/tasks/crafting_profile.json	Add TODO header + formatting.
profiles/tasks/cooking_profile.json	Add TODO header + formatting.
profiles/tasks/construction_profile.json	Add TODO header + formatting.
profiles/ensemble.json	New ensemble-bot profile wiring panel/judge/arbiter settings.
profiles/dragon-slayer.json	New autonomous dragon progression profile with strict rules/examples.
profiles/cloud-persistent.json	New cloud “persistent” ensemble profile for remote 24/7 operation.
profiles/freeguy.json	Add mode scaffolding + hybrid/local variants.
profiles/azure.json	Add mode scaffolding + hybrid/local variants.
profiles/defaults/_default.json	Tighten prompting; include learnings/wiki; adjust defaults (cooldown/night_bed).
profiles/defaults/survival.json	Add TODO header; enable auto_eat/panic_defense.
profiles/defaults/assistant.json	Add TODO header.
profiles/defaults/creative.json	Add TODO header.
profiles/defaults/god_mode.json	Add TODO header.
patches/prismarine-viewer+1.33.0.patch	Patch viewer entity handling to avoid crashes on unknown bones/entities.
patches/@miner-org+mineflayer-baritone+4.5.0.patch	Patch baritone movement to guard missing “above” block.
package.json	Bump version; add deps (baritone, chromadb, discord.js); add lint/husky/lint-staged; overrides hardening.
main.js	Safer JSON parsing; env-var parsing robustness; remote MindServer mode; profile existence checks.
keys.example.json	Add Discord bot token key placeholder.
eslint.config.js	Update globals and rule severities; add overrides for bot/action code.
data/minecraft_wiki.json	Add curated wiki/recipe/mob/item reference data file.
bots/lintTemplate.js	Prefix unused imports to avoid lint noise.
bots/execTemplate.js	Minor formatting fix.
aws/user-data.sh	New EC2 bootstrap for Docker + AWS CLI + base tooling.
aws/backup.sh	New S3 backup script (world + selected bot memory files).
aws/restore.sh	New S3 restore script with safe stop/start of services.
aws/teardown.sh	New AWS teardown script for infra cleanup.
aws/env-toggle.sh	New helper to toggle AWS vs local environment usage.
aws/setup-ollama-proxy.sh	New host-level socat proxy setup for Ollama via Tailscale.
aws/s3-policy.json	New S3 bucket policy template (TLS-only + role-scoped access).
Dockerfile	Improve caching (copy patches); run `npm test`; drop root privileges in image.
Tasks.Dockerfile	Rework tasks image to Node 22 slim + Java 21 + AWS CLI; add non-root user.
FAQ.md	Formatting and minor content cleanup.
CLAUDE.md	New repository guide for Claude Code.
.husky/pre-commit	Add pre-commit hook to run lint-staged.
.gitignore	Add env/secret/AWS/minecraft-data ignores.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-03-01T13:25:07Z

Tasks.Dockerfile

+# ── Application code ───────────────────────────────────────────────────────
 WORKDIR /mindcraft
-COPY ./server_data.zip /mindcraft
-RUN unzip server_data.zip
+RUN git clone --depth 1 https://github.com/mindcraft-bots/mindcraft.git .

-RUN npm install
+COPY ./server_data.zip /mindcraft/
+RUN unzip -q server_data.zip && rm server_data.zip

+RUN npm ci --omit=dev


This image clones https://github.com/mindcraft-bots/mindcraft.git during build, which makes the container depend on external HEAD and can drift from the code in this repository/PR (and breaks reproducibility/offline builds). Consider using COPY . . (or COPY package*.json + npm ci + COPY . .) so the image builds the checked-in source instead of pulling a different repo version.

Copilot · 2026-03-01T13:25:07Z

src/mindcraft/mindcraft.js

+        const settingsPath = `/tmp/mindcraft_${agent_name}_settings.json`;
+        writeFileSync(settingsPath, JSON.stringify(settings));
+
+        const agentProcess = new AgentProcess(agent_name, null, remoteUrl, settingsPath);
+        agentProcess.start(settings.load_memory || false, settings.init_message || null, agentIndex);


createRemoteAgent writes a settings file to /tmp/mindcraft_${agent_name}_settings.json using the profile-provided agent_name verbatim. If a profile name contains path separators or .., this can write outside /tmp (and also collide with other agents). Consider sanitizing agent_name (e.g., allow only [A-Za-z0-9_-], or wrap with path.basename), and/or generate the filename with a random suffix (mkdtemp/UUID) to avoid collisions.

Copilot · 2026-03-01T13:25:08Z

src/agent/conversation.js

+    async endAllConversations() {
        for (const sender in this.convos) {
            this.endConversation(sender);
        }


endAllConversations() is now async, but it doesn't await the per-sender endConversation(sender) calls. This can race with _resumeSelfPrompter() and leave monitor/self-prompter state inconsistent. Update the loop to await each call (or collect promises and await Promise.all).

Copilot · 2026-03-01T13:25:08Z

src/agent/library/world.js

+    // Temporarily disable break/place to test clear-path feasibility
+    const prevBreak = bot.ashfinder.config.breakBlocks;
+    const prevPlace = bot.ashfinder.config.placeBlocks;
+    bot.ashfinder.config.breakBlocks = false;
+    bot.ashfinder.config.placeBlocks = false;


isClearPath() assumes bot.ashfinder is always present and initialized; accessing bot.ashfinder.config will throw if the Baritone plugin failed to load, is disabled, or the bot isn't fully spawned yet. Add a guard (e.g., return false when !bot.ashfinder?.config), or fall back to the previous mineflayer-pathfinder implementation when Baritone isn't available.

Copilot · 2026-03-01T13:25:08Z

src/agent/commands/index.js

 export function blacklistCommands(commands) {
    const unblockable = ['!stop', '!stats', '!inventory', '!goal'];
    for (let command_name of commands) {
        if (unblockable.includes(command_name)){
            console.warn(`Command ${command_name} is unblockable`);
            continue;
        }
+        blockedCommands.add(command_name);
        delete commandMap[command_name];
        delete commandList.find(command => command.name === command_name);
    }


blacklistCommands() tries to remove an entry from commandList via delete commandList.find(...), but delete on the found object does nothing to the array (and find() doesn't return an index). This means blacklisted commands may still appear when iterating commandList elsewhere. Use findIndex + splice, or rebuild commandList/commandMap from a filtered list.

Copilot · 2026-03-01T13:25:08Z

src/agent/vision/vision_interpreter.js

        if (allow_vision) {
-            this.camera = new Camera(agent.bot, this.fp);
+            import("./camera.js").then(({ Camera }) => {
+                try {
+                    this.camera = new Camera(agent.bot, this.fp);
+                    this.camera.on('error', (err) => {
+                        console.warn(`[Vision] Camera async init failed: ${err.message}`);
+                        console.warn('[Vision] Vision disabled — bots will continue without screenshot capability.');
+                        this.allow_vision = false;
+                        if (this.camera) this.camera.destroy();
+                        this.camera = null;
+                    });
+                } catch (err) {
+                    console.warn(`[Vision] Camera init failed (WebGL not available): ${err.message}`);
+                    console.warn('[Vision] Vision disabled — bots will continue without screenshot capability.');
+                    this.allow_vision = false;
+                    this.camera = null;
+                }
+            }).catch((err) => {
+                console.warn(`[Vision] Failed to load camera module: ${err.message}`);
+                console.warn('[Vision] Vision disabled — prismarine-viewer/canvas not available.');
+                this.allow_vision = false;
+                this.camera = null;
+            });


The camera is now initialized asynchronously via dynamic import(), but allow_vision remains true during that async window and this.camera may be undefined when lookAtPlayer() / lookAtPosition() call this.camera.capture(). This can throw at runtime. Consider setting this.camera = null up front and gating on this.camera?.ready (or awaiting a ready promise) before attempting captures, otherwise return a clear "camera not ready" message.

Copilot · 2026-03-01T13:25:08Z

eslint.config.js

+      globals: {
+        ...globals.node,
+        ...globals.browser,
+        // Custom globals for Minecraft bot code
+        skills: "readonly",
+        log: "readonly",
+        world: "readonly",
+        bot: "readonly",
+        agent: "readonly",
+        Vec3: "readonly",
+        newAction: "readonly",
+        nearbyEntities: "readonly",
+        assert: "readonly",
+        chat_model_profile: "readonly",
+        result: "readonly",
+        Compartment: "readonly",
+        res: "writable",
+        id: "readonly",
+        cleanEmb: "readonly",
+        text: "readonly",
+        meta: "readonly",
+        sendRequest: "readonly",
+      },


eslint.config.js declares a number of ordinary identifiers as globals (id, text, meta, result, etc.). These don't look like true runtime globals and will hide real no-undef bugs across the codebase. Prefer scoping these globals to the specific generated bot/action files that need them (via a files override), or remove them and fix the underlying undefined-variable usage.

Copilot · 2026-03-01T13:25:09Z

Tasks.Dockerfile

 WORKDIR /mindcraft
-COPY ./server_data.zip /mindcraft
-RUN unzip server_data.zip
+RUN git clone --depth 1 https://github.com/mindcraft-bots/mindcraft.git .


This Dockerfile step pulls and executes arbitrary code from the unpinned third‑party GitHub repository mindcraft-bots/mindcraft at build time, which is a supply-chain risk. If the upstream repo or the network path is ever compromised, an attacker could inject malicious code into the container image and gain access to any AWS credentials or other secrets used by tasks run in this image. Pin this dependency to a specific immutable identifier (e.g., a commit SHA or release tarball) and, where possible, verify its integrity via checksum or signature instead of cloning the moving default branch.

src/agent/commands/index.js - blacklistCommands: fix delete-on-find bug — delete commandList.find(...) did nothing to the array; replace with findIndex + splice so blacklisted commands are actually removed from commandList (not just commandMap) Tasks.Dockerfile - Replace 'git clone --depth 1 mindcraft-bots/mindcraft' with 'COPY . .' Fixes: broken reproducibility (upstream HEAD drift), offline CI failure, supply-chain risk (arbitrary code fetched at image build time) eslint.config.js - Scope action-code-only globals (newAction, nearbyEntities, assert, result, id, cleanEmb, text, meta) to the action-code files override block instead of declaring them globally, preventing them from masking real no-undef bugs - Keep Compartment, res, sendRequest, chat_model_profile as true globals since they are used in real source files (lockdown.js, item_goal.js, novita.js, prompter.js) src/agent/coder.js (follow-up to security commit) - Spread skills/world into plain objects before Object.freeze() to satisfy ESLint no-import-assign on namespace imports; security property preserved Already fixed in earlier RCs (no change needed): - mindcraft.js: agent_name sanitized with path.basename + regex (RC30) - conversation.js: endAllConversations awaits endConversation (RC30) - world.js: isClearPath guards bot.ashfinder?.config (RC30) - vision_interpreter.js: lookAtPlayer/lookAtPosition guard !this.camera

…DB memory Adds the 3-phase ensemble pipeline used by CloudGrok: - panel.js: queries 4 Grok models in parallel (60s timeout) - arbiter.js: heuristic scoring (length, completeness, action quality, latency); escalates to judge when top-2 within 0.08 margin - judge.js: LLM-as-judge (grok-4) picks best response; 30s timeout with fallback - feedback.js: ChromaDB vector memory (3072-dim Gemini embeddings); retrieves similar past decisions (similarity > 0.6) and injects as [PAST EXPERIENCE] - logger.js: writes every decision to bots/{BotName}/ensemble_log.json - controller.js: EnsembleModel class — drop-in replacement for any single model Integration: profile.model = 'ensemble' routes through EnsembleModel, which implements the same sendRequest() interface as all other model providers.

…tone pathfinding Ender Dragon automation (dragon_runner.js + dragon_progress.js): - 6 gameplay chunks: getDiamondPickaxe, buildNetherPortal, collectBlazeRods, collectEnderPearls, locateStronghold, defeatEnderDragon - Persistent state (dragon_progress.json) with atomic writes and corruption recovery - 5 retries per chunk with exponential backoff; death recovery returns to drop coords - !beatMinecraft / !dragonProgression commands (120-180 min timeout) Baritone A* pathfinding (RC25+): - Custom A* pathfinder replacing mineflayer-pathfinder; distance-adaptive timeouts - Ghost block handling for Paper servers (re-fetch block after nav) - isClearPath() with Baritone integration; null-guard on isWalkable for respawn Skills RC13-RC29 (skills.js + world.js): - RC13: safeToBreak filter fix for tree logs - RC14-17: multi-hop smart explore (200-block relocation), water avoidance, wider collect range - RC18-20: resilient collectBlock — retry on combat, exclude failed positions - RC22: Aikar GC flags; RC23: bypass bot.collectBlock.collect() for Paper - RC24: timeout-protected dig (goToPosition 15s, dig 10s, pickup 8s) - RC26: prefer doors over block-breaking; stale dig fix (re-fetch after nav) - RC27: 9 runtime bug fixes from log review

Vision system (src/agent/vision/): - Dynamic import() for camera module with graceful fallback if WebGL unavailable - Xvfb + Mesa software rendering in Docker (LIBGL_ALWAYS_SOFTWARE=1) - 2-second delay after Xvfb start for WebGL context initialization - Patched prismarine-viewer: entity bone parent null check, unknown entity suppression - Per-bot vision model support (grok-2-vision-1212 for CloudGrok) HUD overlay (mindserver.js + public/index.html): - Gaming-style web dashboard at :8080 - Per-bot panels: runtime tracker, current goal, action display, scrollable command log - Live bot camera feeds via protocol-aware viewer iframes - Toolbar with bot controls; responsive CSS Discord bot (discord-bot.js): - Direct bot chat via Discord channels - Admin commands: !start, !stop, !restart with group-based control - Auto-fix monitor: watches bot-output events for errors, suggests fixes - Role-based access (DISCORD_ADMIN_IDS), usage tracking (!usage [agent|all]) - Path traversal guard + command injection detection on all user input - MindServer integration: live agent status display Windows launcher (start.ps1): one-command start/stop/detach for all bot profiles

…ctions, key loading New security modules: - message_validator.js: injection detection (shell commands, backticks, pipe-to-shell), character sanitization, and length limits for Discord/Minecraft chat - rate_limiter.js: sliding-window per-user limiter with automatic stale entry cleanup - usage_tracker.js: token usage tracking with cost estimation for active providers Existing file hardening: - keys.js: environment variables always override keys.json; added .env support - speak.js: TTS sanitization to prevent command injection in text-to-speech - commands/index.js: isCommandBlocked() check against per-profile blocked_actions array; settings.js deepSanitize() strips __proto__, constructor, prototype from SETTINGS_JSON

…Grafana monitoring Docker / Compose: - docker-compose.aws.yml: production config with LiteLLM proxy (:4000), ChromaDB, Tailscale sidecar, ENFORCE_SECURE_PROFILE=FALSE for mineflayer chat - docker-compose.yml: dev compose with Ollama host routing, port 19565 (was 25565) - Dockerfile: non-root user, Xvfb/Mesa for vision, memory 1536M→2560M - Tasks.Dockerfile: separate image for task evaluation runner - .dockerignore: exclude bot logs, node_modules, world saves from build context AWS scripts (aws/): - ec2-go.sh: one-command deploy (pull/rebuild/restart); IMDSv2 support, auto-detects local vs remote execution - setup.sh: full EC2 provisioning — Docker, Tailscale, ChromaDB, environment setup - deploy.sh: rsync-based deployment with SSM secret pulling - ec2-deploy.sh: self-contained bootstrap for EC2 browser SSH - env-toggle.sh: switch between cloud/local/hybrid environment configs - backup.sh / restore.sh: S3 world backup and restore - setup-ollama-proxy.sh: socat systemd service for Tailscale→Ollama routing Observability: - prometheus-aws.yml: Prometheus scrape config for EC2 deployment - grafana-provisioning/: dashboards, datasources, and alerting rules Security: - whitelist.json: pre-built offline UUIDs (avoids Playerdb crash for ONLINE_MODE=FALSE) - .env.example / keys.example.json: API key templates - .husky/pre-commit: ESLint zero-warning gate on every commit Patches: - prismarine-viewer: entity bone parent null check, unknown entity suppression

…ation, ESLint, and CI/CD New profiles: - ensemble.json: all-Grok 4-panel (grok-4, grok-4-fast-non-reasoning, grok-4-1-fast-non-reasoning, grok-code-fast-1); judge: grok-4 (30s) - cloud-persistent.json: grok-4 single-model cloud bot - dragon-slayer.json: sweaterdog/andy-4:q8_0 via Ollama; autonomous dragon speedrun - local-research.json: sweaterdog/andy-4 for research and exploration - All existing profiles updated: blocked_actions, conversation_examples, wiki references Core agent improvements: - agent.js: death handler + respawn recovery, post-respawn grace period (5s), human-player priority, max-command cap at 15, empty init_message support - action_manager.js: action deadlock detection, cancelResume(), error propagation - modes.js: night bed mode, door navigation, anti-team-kill guards, food panic fix - history.js / learnings.js: atomic writes (.tmp + rename), EBADF retry with backoff - conversation.js: inter-bot messaging protocol; Grok_En alias system - memory_bank.js, self_prompter.js, coder.js, connection_handler.js: various fixes - agent_process.js: 3 quick-exit retries with increasing delay Model providers (src/models/): - prompter.js: \, \, \, \, \ placeholder injection - grok.js: updated for grok-4 / grok-4-fast-non-reasoning API surface - ollama.js: fix http.request (avoid headers timeout), num_ctx/num_gpu params - All other providers: price table updates, API compatibility fixes Data / wiki: - data/minecraft_wiki.json: Minecraft wiki data for \ prompt injection Configuration and tooling: - settings.js: deepSanitize() prototype-pollution guard, SETTINGS_JSON env override, allow_vision config, port 19565 (was 25565) - main.js: multi-profile startup cleanup - package.json: updated deps + patch-package postinstall - eslint.config.js: flat config migration; zero-warning enforcement - CLAUDE.md: AI assistant guidance for codebase navigation - FAQ.md, README.md: updated for v0.1.3 Hybrid Research Rig

Z0mb13V1 · 2026-03-02T11:29:28Z

Replaced by a fresh PR to clear edit history. All content is identical — see the new PR for a clean description.

Copilot AI review requested due to automatic review settings March 1, 2026 13:18

Copilot started reviewing on behalf of Z0mb13V1 March 1, 2026 13:19 View session

Copilot AI reviewed Mar 1, 2026

View reviewed changes

Z0mb13V1 force-pushed the pr/hybrid-rig-v0.1.3 branch from e482dbc to 76f10a7 Compare March 1, 2026 14:10

Z0mb13V1 added 6 commits March 2, 2026 05:57

Z0mb13V1 force-pushed the pr/hybrid-rig-v0.1.3 branch from 3404c87 to 88213bc Compare March 2, 2026 11:05

security: remove hardcoded EC2 IP from all public-facing files

fe74365

Z0mb13V1 closed this Mar 2, 2026

Z0mb13V1 reopened this Mar 2, 2026

Z0mb13V1 closed this Mar 2, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Hybrid Research Rig v0.1.3 — ensemble bot, Baritone pathfinding, vision, HUD, security hardening#714

Hybrid Research Rig v0.1.3 — ensemble bot, Baritone pathfinding, vision, HUD, security hardening#714
Z0mb13V1 wants to merge 7 commits intomindcraft-bots:developfrom
Z0mb13V1:pr/hybrid-rig-v0.1.3

Z0mb13V1 commented Mar 1, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Copilot AI Mar 1, 2026

Uh oh!

Z0mb13V1 commented Mar 2, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

Z0mb13V1 commented Mar 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 1, 2026

Choose a reason for hiding this comment

Uh oh!

Z0mb13V1 commented Mar 2, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Z0mb13V1 commented Mar 1, 2026 •

edited

Loading