@rohitg00

write-agentmemory-skill defines the canonical format and governance rules for agentmemory skills, ensuring consistency, scannability, and maintainability across the cross-vendor skill graph.

iii-agentic-backend coordinates multi-agent pipelines using named queues, shared state, and approval gates. agents register as functions, hand work via async queues, and broadcast completion events through pubsub.

iii-functions-triggers documents function registration and event binding across typescript, python, and rust on the iii engine. covers synchronous invocation, durable queues, and cross-language function composition with auto-generated request/response schemas.

recall retrieves past observations and learnings from agent memory using hybrid search, surfacing high-signal results grouped by session without fabrication or invention.

iii-reactive-backend provides event-driven primitives for real-time state management and client streaming on the iii engine. covers state crud, automatic change triggers, and websocket broadcast patterns with a reference todo app.

handoff retrieves the most recent agent session for the current working directory and surfaces any unanswered question, with directory-boundary matching to avoid sibling repo confusion.

session-history retrieves and formats recent project sessions from memory in reverse chronological order, showing session metadata, observation counts, and key decisions without fabrication.

agentmemory-architecture describes the iii-engine design of a local memory server for agents, covering hybrid retrieval, storage lifecycle, and network ports, but lacks procedural depth and actionable guidance.

recap retrieves and groups recent agent sessions by date, extracting high-importance observations. execution depends on memory service availability with mcp-first fallback to http.

agentmemory-config covers environment setup, feature flags, api key injection, and auth via ~/.agentmemory/.env file. operates in zero-llm mode by default and requires server restart after config changes.

remember persists user insights and decisions to long-term storage with concept tagging for retrieval, accepting freeform content plus optional file references.

k8s-multicluster covers context switching, parallel cluster queries, helm deployments, and cluster lifecycle management via capi. lacks actionable decision logic and failure recovery patterns.

agentmemory-agents documents how to wire the memory server into coding agents via connect command, with fallback to claude-code and post-connection verification steps.

agentmemory-mcp-tools provides a reference catalog of memory system operations grouped by function (capture, retrieve, sessions, governance) without actionable guidance on when to use each tool or how its parameters relate to user intent.

forget removes selected observations from agentmemory via search-then-delete workflow. requires explicit user confirmation before deletion and logs reason for audit trail.

commit-history retrieves recent git commits associated with agent sessions, with optional filtering by branch and repository. outputs formatted commit metadata linked to session ids.

commit-context maps code locations to the agent session and commit that introduced them, using git blame output and memory lookup to surface session metadata and observation counts.

k8s-browser wraps playwright-style automation for kubernetes web interfaces. covers dashboard, grafana, and argocd ui navigation via selectors and waits, but lacks intent clarity and failure handling.

The agentmemory plugin hooks that capture observations automatically across the agent session lifecycle. Use when explaining how memory gets captured without…

The agentmemory HTTP REST API surface, the primary protocol for talking to the memory server. Use when calling agentmemory over HTTP, when MCP is unavailable…

Cluster API lifecycle management for provisioning, scaling, and upgrading Kubernetes clusters. Use when managing cluster infrastructure or multi-cluster operations.

Kubernetes backup and restore with Velero. Use when creating backups, restoring applications, managing disaster recovery, or migrating workloads between clusters.