Todo for AI MCP Server MCP Policy · 244 Tools

POLICY

How to control Todo for AI MCP Server

PolicyLayer is an MCP gateway — it sits between your AI agents and Todo for AI MCP Server, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "cancel_workflow_run": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "add_deliberation_message": {
    "limits": [
      {
        "counter": "add_deliberation_message_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "acknowledge_conflict": {
    "limits": [
      {
        "counter": "acknowledge_conflict_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register Todo for AI MCP Server — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON TODO FOR AI →

Instant setup, no code required.

FULL CATALOGUE

All 244 Todo for AI MCP Server tools

DESTRUCTIVE 11 tools

Destructive cancel_workflow_run Cancel a running workflow. All pending/running steps are cancelled. Destructive clear_step_runtime_override Clear runtime overrides for a workflow step run, reverting to the workflow definition. Destructive delete_agent_experience Delete (soft-delete) an experience record for an Agent. Destructive delete_collaboration_template Delete a user-created collaboration template. Destructive delete_knowledge_entry Delete (invalidate) a knowledge entry. Destructive delete_sandbox Delete a sandbox policy (only if no active executions reference it). Destructive delete_shared_context Delete a shared context entry by its ID. Destructive delete_workflow Delete a workflow definition. Destructive delete_workflow_trigger Delete a workflow trigger. Destructive revoke_cross_project_agent Revoke an Agent\ Destructive revoke_sandbox_execution Manually revoke (terminate) a running sandboxed execution.

EXECUTE 7 tools

Execute append_run_logs Append log entries to a specific Agent run. Use this during execution to record progress, tool calls, decision Execute escalate_overdue_tasks Trigger priority auto-escalation for overdue tasks. Tasks past their due date that are not yet urgent will hav Execute handoff_task Hand off a task from its current Agent to another Agent. Cancels the active assignment (if any) and creates a Execute launch_workflow Launch a new run of a workflow. Creates tasks for each step, assigns matching Agents, and starts entry-point s Execute pause_workflow_run Pause a running workflow. Already-running steps continue, but no new steps will be started until resumed. Execute retry_workflow_run Retry a failed workflow by resetting failed/skipped steps and re-advancing the DAG. Execute start_sandbox_execution Start a new sandboxed execution for an Agent. Freezes the policy snapshot and returns the execution record + p

WRITE 47 tools

Write add_deliberation_message Add a deliberation message (argument/evidence/comment) to a deliberation protocol. Enables multi-round discuss Write apply_capability_adaptation Apply capability adaptations to an Agent (add or remove capabilities). Write apply_experience_decay Apply time-based confidence decay to an Agent\ Write auto_resolve_conflicts Maintenance: scan for conflicts and auto-resolve low-severity ones (INFO/WARNING) using their suggested strate Write broadcast_message Send a broadcast message from one Agent to all other active Agents owned by the same user. Useful for coordina Write complete_sandbox_execution Mark a sandboxed execution as completed with aggregated usage stats. Write complete_workflow_step Mark a workflow step as completed (or failed) and advance the DAG. This is the callback that Agents call when Write create_agent Create an Agent identity and declare its collaboration capabilities Write create_agent_experience Manually create an experience record for an Agent. Write create_channel Create a collaboration channel for multi-Agent discussion. Agents can be auto-added as members. Write create_collaboration_template Create a custom collaboration template defining a team of Agents. Write create_knowledge_entry Create a knowledge entry for an Agent. Store insights, patterns, or solutions for future reuse. Write create_protocol Create a collaboration protocol (proposal, vote, consensus, auction, or handoff) for structured multi-Agent de Write create_sandbox Create a new Agent sandbox policy governing tool/network/filesystem/resource access during execution. Write create_subtask Create a child task under a parent task so an Agent can decompose work. The child inherits the parent\ Write create_task Create a new task in the specified project Write create_task_template Create a reusable task template. Templates let you quickly create new tasks with predefined title, content, pr Write create_workflow Create a new workflow definition with steps. Each step specifies required Agent capabilities, optional task te Write create_workflow_trigger Create a scheduled trigger for a workflow. Supports cron expressions for recurring runs and one-shot schedules Write export_security_events Export the unified security event feed as CSV or JSON text (up to 1000 rows). Accepts the same filters as list Write heartbeat_agent Record a heartbeat for an Agent and optionally update its availability status Write instantiate_sandbox_template Create a sandbox policy from a preset template, optionally overriding fields and binding to an agent. Write instantiate_task_template Create a new task from a template. The task inherits the template\ Write instantiate_workflow_template Create a workflow from a built-in template. Optionally override name and project context. Write mark_notifications_read Mark one or more notifications as read. Pass all=true to mark everything read, or pass an array of IDs to mark Write mark_offline_agents Scan all active/paused agents and mark those whose last heartbeat exceeds the offline threshold (30 min) as OF Write post_task_event Post a collaboration message to a task timeline so Agents can communicate: leave notes, ask/answer questions, Write register_capabilities Register new capabilities for an Agent at runtime. Use mode Write resolve_conflict Resolve a conflict with a chosen strategy. For duplicate claims, FIRST_WINS/HIGHEST_REPUTATION/LEAST_LOADED al Write resolve_protocol Manually resolve a protocol (force accept/reject/cancel). Write resume_workflow_run Resume a paused workflow. The DAG engine re-evaluates which steps can start. Write self_register_agent Self-register an Agent into the platform. If an agent with the same name+provider already exists, it updates t Write send_agent_message Send a direct message from one Agent to another. Delivered via SSE (real-time) and Notification (persistent). Write send_channel_message Send a message to a collaboration channel. All channel members will receive the message. Write set_shared_context Create or update a shared context entry for a task (upsert by task_id + key). Use this to store intermediate r Write set_step_runtime_override Dynamically reconfigure a not-yet-terminal workflow step (runtime overrides). Only PENDING/WAITING steps accep Write set_task_dod Set a task Write submit_task_feedback Submit feedback for a completed or in-progress task Write update_agent Update an Agent identity, status, model metadata, or capabilities Write update_agent_assignment Update an Agent assignment state, progress, feedback, or execution result Write update_agent_experience Update an experience record for an Agent. Write update_knowledge_entry Update a knowledge entry (title, content, tags, confidence, etc.). Write update_sandbox Update an existing sandbox policy. Write update_task_assignment Update a task assignment as the current user or coordinator. Use this to approve, resume, or cancel items from Write update_task_status Update a task Write update_workflow Update a workflow definition and its steps (full replacement of steps if provided). Write update_workflow_trigger Update a workflow trigger. Only provided fields are changed.

READ 179 tools

Read acknowledge_conflict Mark a conflict as acknowledged (seen). Read authorize_cross_project_agent Authorize an Agent to work in a different project. Requires ADMIN/OWNER access to the target project. Read auto_extract_experiences Auto-extract experiences from an Agent\ Read auto_extract_knowledge Auto-extract knowledge from an Agent\ Read bind_agent_sandbox Bind a sandbox policy to an Agent (replaces any existing active binding). Read check_sandbox_action Dry-run check whether an action (tool/network/fs_write) is permitted under a sandbox policy. Read claim_agent_task Claim a specific task or the next claimable task for an Agent Read claim_cross_project_task Claim a task from a cross-project for an Agent. Requires cross-project authorization. Read collaboration_graph Platform-wide Agent collaboration graph from direct-message audit logs — nodes (agents with message totals) an Read collaboration_metrics Retrieve aggregate collaboration metrics for the dashboard: task completion rates, agent utilization, workflow Read decay_all_experiences System maintenance: apply decay to all agents\ Read diff_workflow_versions Compare two versions of a workflow. Returns added, removed, and modified steps. Read discover_agents Find available Agents by capability, role, or kind. Useful for discovering which agents can handle a specific Read discover_cross_project_agents Discover agents across all accessible projects, optionally filtered by capability. Read dispatch_tasks Coordinator auto-dispatch: have a coordinator Agent distribute claimable (unassigned) tasks across the owner\ Read find_capable_agents_cross_project Find agents across all accessible projects that have specific capabilities. Read find_cross_project_tasks Find claimable tasks across all projects an Agent is authorized for. Read fire_due_triggers Check all active workflow triggers and fire those that are due. Designed to be called periodically by an exter Read get_agent_capability_gap_analysis Agent capability gap analysis. Compares each agent\ Read get_agent_capability_supply_demand Analyze supply vs demand for each capability. Supply = agents declaring the capability; demand = active tasks Read get_agent_collaborators Aggregate an Agent\ Read get_agent_cross_project_efficiency Measure realized value of cross-project Agent authorizations. For each authorization into a project owned by t Read get_agent_experience Get a specific experience record for an Agent. Read get_agent_experiences_decay_alerts Flag Agents whose experience-base confidence is declining. Compares older-half vs newer-half average confidenc Read get_agent_failure_error_patterns Agent failure error pattern clustering. Groups FAILED AgentRun rows by error text prefix, returning pattern cl Read get_agent_failure_reasons Distribution of Agent run failure reasons. Buckets FAILED AgentRun rows by a normalized error type (first line Read get_agent_health Per-Agent composite health score (0-100) for the current user, combining reputation (0.4), assignment completi Read get_agent_health_alerts Low-health Agent alert list for the current user. Returns Agents whose composite health_score < min_health_sco Read get_agent_health_state_transitions Agent health state transition flow. Classifies each agent-day as healthy/degraded/critical based on reputation Read get_agent_health_trend Daily reputation-derived health trend for the current user Agents. Aggregates reputation.update audit entries Read get_agent_idle_ranking Rank Agents by idle duration. Idle time is measured from the most recent of last_seen_at and the latest TaskAs Read get_agent_inbox Retrieve collaboration messages directed at a specific Agent (its @mention inbox) across all of the owner\ Read get_agent_messages Get recent direct messages for an Agent (both sent and received). Read get_agent_productivity Per-Agent productivity stats for the current user: total assignments, done, failed, cancelled, expired, in-pro Read get_agent_productivity_alerts Low-efficiency Agent alert list for the current user. Returns Agents whose completion rate < min_completion_ra Read get_agent_productivity_by_kind Productivity comparison grouped by Agent kind for the current user. For each kind (assistant/worker/orchestrat Read get_agent_productivity_calendar_heatmap Date × Agent completion calendar heatmap. Buckets done assignments by calendar date (YYYY-MM-DD) and agent_id Read get_agent_productivity_hourly_heatmap Hour-of-day × Agent completion heatmap. Buckets done assignments by the hour (0-23) of completed_at and agent_ Read get_agent_productivity_trend Daily Agent assignment completion trend for the current user. Per-day done and failed TaskAssignment counts wi Read get_agent_productivity_weekly_comparison Week-over-week Agent productivity comparison. Returns per-agent done counts for current week vs previous week Read get_agent_reputation Get the reputation record for an Agent. Reputation affects task assignment priority. Read get_agent_reputation_history Get the timeline of notable reputation changes for an Agent (failures and quality-feedback deltas), reconstruc Read get_agent_run_resource_trend Per-agent daily run count and average duration trend. Returns sparkline-friendly daily series for run count an Read get_agent_run_resource_usage Agent run resource usage ranking. Per-agent: total runs, total wall-clock hours, average run duration in minut Read get_agent_sandbox Get the active sandbox bound to an Agent (if any). Read get_agent_skill_matching Agent skill matching recommendation. For unassigned in-progress tasks, match task keywords to agent capabiliti Read get_agent_specialization_evolution Track how each Agent domain coverage evolves over time. Weekly distinct domain count series, revealing special Read get_agent_task_handoff_stats Agent task handoff statistics. Aggregates handoff events by (from_agent, to_agent) pairs, counts frequency and Read get_agent_workload_forecast Forecast each Agent near-future task load via linear regression on daily assignment counts. Returns per-agent Read get_channel_activity_trend Channel activity trend. Per-channel daily message count sparkline and active member count over the lookback wi Read get_collaboration_graph_timeline Day-by-day collaboration graph snapshots for timeline replay. Returns bucketed (day/week) snapshots of collabo Read get_conflict Get a conflict record by ID. Read get_conflicts_by_agent Per-Agent conflict involvement counts (total + active), top N by total, enriched with name/kind. Reveals which Read get_conflicts_dashboard Aggregate conflict stats for the current user: totals, active count, breakdowns by type/status/severity. Read get_conflicts_sandbox_correlation Cross-dimension correlation between Agent conflicts and sandbox violations. For each conflict, checks whether Read get_conflicts_strategy_stats Resolution strategy effectiveness: per-strategy usage count and recurrence rate (fraction of resolved conflict Read get_conflicts_trend Daily conflict detection vs resolution counts over the last N days (default 30). Each bucket has {date, detect Read get_experience_validation_stats Get validation statistics for an Agent\ Read get_experiences_confidence_decay_forecast Confidence decay forecast using linear regression on daily averages. Projects 7 days ahead, reports regression Read get_experiences_confidence_distribution Confidence interval distribution for the user valid experiences. Buckets into 5 intervals (0-0.2, 0.2-0.4, 0.4 Read get_experiences_decay_by_domain Per-domain decay comparison for the user valid experiences. Aggregates by domain: total, active (confidence>=0 Read get_experiences_decay_by_task_type Per-task-type decay comparison for the user valid experiences. Same as decay-by-domain but grouped by task_typ Read get_experiences_low_confidence List the current user valid experiences with confidence below max_confidence (default 0.5), sorted by confiden Read get_experiences_propagation_chain Experience sharing propagation chain. Groups shared experiences by source agent with shared_count, total_reuse Read get_experiences_reuse_trend Daily reuse + decay trend for the user valid experiences. Buckets experiences by last-reused (or creation) dat Read get_experiences_scatter Confidence × reuse-count scatter points for the user valid experiences. One point per experience: confidence, Read get_experiences_skill_coverage_radar Per-Agent skill coverage radar across experience domains. Returns top N domains and per-agent normalized score Read get_experiences_source_distribution Experience count by creation source. Groups by origin: manual (no workflow run), workflow (has source_workflow Read get_experiences_stats Aggregate AgentExperience stats for the current user: totals, breakdowns by domain/task_type/experience_type, Read get_knowledge_entry Get a specific knowledge entry by ID. Read get_knowledge_propagation_network Cross-Agent knowledge propagation network. Nodes are Agents sharing experiences, edges connect contributors we Read get_orchestrator_status Return the state of the built-in orchestrator scheduler (enabled/disabled) and the last orchestration cycle su Read get_project_info Get detailed project information including statistics and configuration. Provide either project_id or project_ Read get_project_tasks_by_name Get all pending tasks for a project by project name, sorted by creation time Read get_protocol Get a protocol with all its messages/responses. Read get_protocol_analytics Get analytics for collaboration protocols: usage by type, resolution rates, participation. Read get_protocol_decision_latency Collaboration protocol decision latency analysis. For resolved protocols, computes creation-to-resolution late Read get_run_logs Retrieve append-only log entries for a specific Agent run. Agents write logs during execution to record progre Read get_sandbox Get a sandbox policy by ID. Read get_sandbox_dashboard Get aggregate sandbox stats for the current user: total sandboxes, executions, violations, breakdowns. Read get_sandbox_execution Get a sandbox execution record with its violations. Read get_sandbox_template_usage Sandbox policy template instantiation stats: per-template usage count and how many instances were bound to an Read get_sandbox_violation_trend Daily sandbox violation counts + by-type breakdown over the last N days (default 30). Useful for spotting whet Read get_sandbox_violations_by_agent Per-Agent sandbox violation counts over the last N days, top N by total, enriched with name/kind and a by-viol Read get_shared_context Read shared context entries for a task. Agents use the shared context to persist intermediate results, referen Read get_step_effective_params Get the effective parameters for a workflow step run (runtime overrides merged with the step definition). Read get_step_sandbox_execution Get the sandboxed execution (if any) bound to a workflow step run. Surfaces the auto-started execution + froze Read get_task_allocation_fairness Task allocation fairness analysis. Computes Gini coefficient and Lorenz curve for task distribution across age Read get_task_by_id Get detailed task information by task ID Read get_task_comment_sentiment_trend Task comment sentiment trend. Aggregates comment events by day and classifies sentiment (positive/negative/neu Read get_task_completion_by_assignee Daily task completion trend grouped by assignee (Agent). Buckets done assignments by calendar day of completed Read get_task_completion_by_priority Task completion rate by priority. Groups tasks by priority with total/done/cancelled/completion_rate. Reveals Read get_task_completion_by_project Daily task completion trend grouped by project. Buckets done tasks by calendar day of completed_at and project Read get_task_completion_forecast Task completion forecast based on historical velocity. Computes daily done-task velocity over the lookback win Read get_task_completion_rate_by_project Task completion rate snapshot comparison across projects. Groups tasks by project with total/done/in_progress/ Read get_task_dependency_chain Task dependency chain analysis. Finds root tasks with subtask hierarchies, computes chain depth, total tasks, Read get_task_evidence Get a task Read get_task_overdue_by_assignee Overdue task count grouped by assignee (Agent). Per agent: overdue count, by-priority breakdown, earliest over Read get_task_overdue_clustering Overdue task clustering analysis by project and priority. Per cluster: project name, priority, count, avg days Read get_task_overdue_trend Daily overdue task trend by due_date: per-day overdue count (due_date<now, status not done/cancelled) plus per Read get_task_priority_trend Daily task priority distribution trend. Groups tasks by created_at date and priority level (critical/high/medi Read get_task_rework_analysis Task rework analysis. Finds tasks reverted from done/review back to in_progress/todo, counts per-task rework, Read get_task_stats Aggregate task lifecycle stats for the current user projects: total, by_status, by_priority, completion/cancel Read get_workflow Get a single workflow definition with its steps. Read get_workflow_failed_steps_by_duration Rank failed workflow steps by average wall-clock duration (finished - started). Per step_key: failures count, Read get_workflow_failure_correlation Cross-dimension correlation between failed workflow steps and collaboration conflicts / sandbox violations. Fo Read get_workflow_failure_correlation_by_step Per-step-key failure correlation with conflicts / sandbox violations. Aggregates the same ±window_hours co-occ Read get_workflow_run Get a single workflow run with step details. Read get_workflow_run_console Step-level real-time console for a workflow run. Aggregates each step run with its sandbox execution, effectiv Read get_workflow_run_duration_percentiles Daily trend of workflow run duration percentiles (P50/P90/P95). For each day, aggregates completed WorkflowRun Read get_workflow_run_trend Daily workflow run outcome trend: per-day succeeded vs failed counts over the last N days (default 30), using Read get_workflow_similarity_matrix Workflow run similarity matrix. Computes pairwise Jaccard similarity between workflow run step_key sets. Retur Read get_workflow_step_bottleneck_timeline Per-step daily average duration timeline. Tracks how each workflow step avg duration changes over time to spot Read get_workflow_step_cofailure_matrix Step-key co-failure matrix. For each failed workflow run, collects failed step_keys. Builds a symmetric co-occ Read get_workflow_step_dependency_bottleneck Workflow step dependency bottleneck analysis. Computes DAG critical path (longest duration path) per workflow, Read get_workflow_step_duration_histogram Workflow step duration histogram. Buckets completed step durations into time ranges (0-10s, 10-30s, 30-60s, 60 Read get_workflow_step_failure_rate Per-step-key failure rate ranking. For each step_key: total runs, failed count, failure rate percentage. Sorte Read get_workflow_step_hourly_distribution Step execution hour-of-day distribution. Groups WorkflowStepRun by (step_key, hour) based on started_at. Retur Read get_workflow_step_retry_topology Step retry topology. Groups WorkflowStepRun by step_key, counting retries (attempt>1), first-attempt success r Read get_workflow_step_stats Per-step-key execution stats across the user workflow runs: total/succeeded/failed/skipped, success rate, and Read get_workflow_structural_complexity Analyze design-time structural complexity of active workflows. For each workflow computes DAG metrics from ste Read get_workflow_success_rate_by_workflow Per-workflow run success rate comparison over the last N days (default 30). Returns top workflows by total fin Read get_workflow_template Get details of a specific workflow template by key. Read get_workflow_version Get a specific version snapshot of a workflow. Read health_check Run a full platform health check: expire stale agents, expire stale leases, and escalate overdue tasks. Design Read ignore_conflict Dismiss a conflict without action. Read instantiate_collaboration_template Instantiate a collaboration template: creates the Agents, a collaboration channel, and optionally launches the Read learn_from_experience An Agent internalizes a shared experience from another Agent for collective learning. Read list_agent_assignments List task assignments for an Agent Read list_agent_cross_projects List all projects an Agent is authorized to work in. Read list_agent_experiences List experiences for an Agent. Experiences capture success/failure patterns from past tasks. Read list_agents List Agent identities available to the current user Read list_audit_logs Query the immutable audit trail for platform operations. Supports filtering by action type, resource type, act Read list_channel_messages List messages in a collaboration channel, in chronological order. Read list_channels List collaboration channels. Multiple Agents can join a channel to discuss and coordinate on tasks. Read list_collaboration_templates List collaboration templates (built-in and user-created). Each template defines a team of Agents that work tog Read list_conflicts List collaboration conflicts for the current user, optionally filtered by status/type. Read list_knowledge_entries List knowledge entries for an Agent. Agents store insights, patterns, and solutions for cross-task reuse. Read list_my_tasks List tasks relevant to you (created by, owned by, in your projects, or assigned to you). Entry point to discov Read list_notifications List persistent notifications for the current user. Notifications are created when collaboration events (claim Read list_orchestrator_history Return recent orchestration run records for trend analysis — per-run stats (stale agents, timed-out steps, tri Read list_project_external_agents List all external Agents authorized to work in a project. Read list_protocols List collaboration protocols (proposals, votes, consensus, auctions, handoffs) with optional filters. Read list_recommended_tasks List unassigned tasks that match an Agent\ Read list_reputations List reputation records for all your Agents, ranked by score. Read list_review_queue List Agent assignments that need human feedback or final review Read list_sandbox_executions List executions under a sandbox policy. Read list_sandbox_templates List preset sandbox policy templates (read_only_research, code_generation, data_analysis, full_autonomy, sandb Read list_sandboxes List Agent sandbox policies owned by the current user. Optionally filter by agent_id or active_only. Read list_security_events Unified security event feed: aggregates sandbox violations, agent conflicts, and security-relevant audit entri Read list_shared_experiences List shared experiences from other Agents that this Agent can learn from. Read list_shared_knowledge List knowledge entries shared with projects the current user is a member of. Read list_task_assignments List Agent assignments for a task, including active leases and assigned Agents Read list_task_events List collaboration events for a task so Agents can inspect handoffs, claims, review requests, and lease expira Read list_task_templates List reusable task templates owned by the current user. Templates define default title, content, priority, tag Read list_workflow_runs List workflow runs for the current user. Read list_workflow_templates List built-in workflow templates for common multi-Agent collaboration patterns (e.g. code review pipeline, res Read list_workflow_triggers List scheduled workflow triggers owned by the current user. Read list_workflow_versions List version history for a workflow. Each update creates a snapshot. Read list_workflows List workflow definitions. A workflow is a DAG of steps that coordinate multiple Agents in sequence or paralle Read orchestrate Global collaboration orchestrator: runs the full multi-Agent maintenance cycle in one call — (1) health (stale Read orchestrator_daily_trend Daily aggregation of orchestration runs, aligned with the security events daily-trend time dimension so the tw Read recalculate_reputation Recalculate an Agent\ Read recommend_experiences Recommend relevant experiences for an Agent given a task context. Searches own and shared experiences. Read report_progress Append a progress log entry to a task (append-only task log) to keep humans informed while working Read report_sandbox_violation Report a policy violation during a sandboxed execution. Optionally terminate the execution. Read report_step_sandbox_violation Report a sandbox policy violation for a workflow step execution. If terminate_step is true, the step is marked Read request_approval Ask the human to decide something about a task (destructive operation, budget, scope change). The request ente Read respond_to_protocol Respond to a protocol (vote, bid, accept, reject, counter-proposal, comment). Some responses auto-resolve the Read rollback_workflow Rollback a workflow to a specific version. Creates a snapshot of the current version first. Read scan_conflicts Run a conflict detection scan: duplicate claims, stale assignments, protocol deadlocks. Creates conflict recor Read search_knowledge Search an Agent\ Read search_tasks Search accessible tasks by keyword in title or content Read security_events_by_agent Per-agent aggregation of security events for ranking. Reuses the same filters as list_security_events. Returns Read security_events_daily_trend Daily aggregation of security events for trend visualization. Reuses the same filters as list_security_events Read share_agent_experience Share an Agent\ Read suggest_capability_adaptation Suggest capability adaptations for an Agent based on its experience patterns. Returns suggested additions/remo Read timeout_workflow_steps Scan all running workflow steps and mark those that have exceeded their timeout_seconds as FAILED, then advanc Read validate_experience Cross-validate an experience by another Agent. Confirm or refute accuracy to adjust confidence.

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about Todo for AI MCP Server

Can an AI agent delete data through the Todo for AI MCP Server MCP server? +

Yes. The Todo for AI MCP Server server exposes 11 destructive tools including cancel_workflow_run, clear_step_runtime_override, delete_agent_experience. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Todo for AI MCP Server? +

The Todo for AI MCP Server server has 47 write tools including add_deliberation_message, apply_capability_adaptation, apply_experience_decay. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Todo for AI MCP Server.

How many tools does the Todo for AI MCP Server MCP server expose? +

244 tools across 4 categories: Destructive, Execute, Read, Write. 179 are read-only. 65 can modify, create, or delete data.

How do I enforce a policy on Todo for AI MCP Server? +

Register the Todo for AI MCP Server MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Todo for AI MCP Server

What Todo for AI MCP Server exposes to your agents

The most dangerous Todo for AI MCP Server tools

How to control Todo for AI MCP Server

All 244 Todo for AI MCP Server tools

Related servers

Questions about Todo for AI MCP Server

Enforce policy on every Todo for AI MCP Server tool call.