Developer Tools●Mid
Git Web Manager
Laravel-specific deployment dashboard when Coolify already does this.
Cozy
wallabydesigns
313mo ago
Back to browse
GitHub Repository
0 starsJupyter Notebook
Akili – gated skill updates and rollback receipts for deployed models
by Akili28·Jul 22, 2026·1 point·0 comments
AI Analysis
●●SolidBig BrainNiche Gem
Validated MuJoCo benchmarks show 99.5% safe skill certification with zero downtime rollbacks.
Strengths
- •Tamper-evident audit chains track every candidate capability from experience to activation.
- •Pluggable engine architecture supports both robotics adapters and LLM agent evolution tracks.
- •Matched-feature CIFAR-100 runs show significantly lower forgetting rates than DER++ baselines.
Weaknesses
- •Raw comparison JSON for agent evolution benchmark still pending import before v0.1.0 release.
- •Independent Mammoth replication of CIFAR results not yet completed or published.
Category
Target Audience
ML engineers building continual learning systems
Similar To
Avalanche · Mammoth · Continual AI
Similar Projects
Developer Tools●Mid
Skrun – Deploy any Agent Skill as an API (open source)
Yet another agent deployment wrapper in a crowded category.
Ship It
frizull
62123mo ago
Developer Tools●Mid
Skrun – Deploy any Agent Skill as an API (open source)
Yet another agent deployment layer when LangChain and LlamaIndex exist.
Ship It
frizull
303mo ago
AI/ML●●●Banger
Map – Receipts and rollback for AI agents
Hash-chained ledgers with COMPENSATE/RESTORE rollback schemas solve agent liability gaps.
Big BrainZero to OneBold Bet
Dahvay
203mo ago
Developer Tools●●Solid
OtaKit – open-source, self-hostable OTA updates for Capacitor apps
Self-hosted CodePush alternative with automatic rollback on failed updates.
Ship ItSolve My Problem
garymiklos
1020d ago
Security●●Solid
Agent Gate – Execution authority for AI agents, vault-backed rollback
The idea of inserting a deterministic 'gate' between proposed tool calls and execution is smart and practical: precomputed classification plus an agent-unreachable vault means destructive operations become reversible by default. The README calls out clear mechanics (envelope checks, vault backup, tiered responses and structured denials) which suggests this is more than a thought experiment — it's a focused infra piece for teams that actually let agents touch production.
Big BrainNiche Gem
hammer32
115mo ago