@zack/incident-response-loop

Portable incident-response loop for assess, execute, review, and revision-driven operational follow-through.

Install
agentpm install @zack/incident-response-loop@0.1.0
Load
import { loadLoop } from '@agentpm/sdk';
const loop = await loadLoop('@zack/incident-response-loop@0.1.0');
Weekly installs
2
0%
Last publish
Today
v0.1.0
agent.json
{
  "name": "incident-response-loop",
  "version": "0.1.0",
  "description": "Portable incident-response loop for assess, execute, review, and revision-driven operational follow-through.",
  "loop": {
    "archetype": "incident-response",
    "entry_phase": "assess",
    "limits": {
      "max_steps": 12
    },
    "phases": [
      {
        "id": "assess",
        "access": {
          "tools": true,
          "memory": {
            "read": true
          },
          "knowledge": true
        },
        "outcomes": [
          {
            "id": "proceed",
            "description": "The incident has enough context to move into active execution."
          },
          {
            "id": "handoff",
            "description": "The incident should be handed off immediately because ownership, authority, or context is insufficient."
          }
        ],
        "objective": "Review the current incident state, identify the active owner and primary risks, and decide whether work should proceed or hand off immediately."
      },
      {
        "id": "execute",
        "access": {
          "tools": true,
          "memory": {
            "read": true,
            "write": true
          },
          "knowledge": true
        },
        "objective": "Carry out the current operational work, update the working state, and prepare the evidence needed for the next review pass."
      },
      {
        "id": "review",
        "access": {
          "tools": true,
          "memory": {
            "read": true,
            "write": true
          },
          "knowledge": true
        },
        "outcomes": [
          {
            "id": "complete",
            "description": "The incident is resolved enough to close the loop cleanly."
          },
          {
            "id": "needs-more-work",
            "description": "The current result is incomplete and another execution pass is required."
          },
          {
            "id": "handoff",
            "description": "The work should hand off because progress is blocked, ownership changed, or escalation is required."
          }
        ],
        "objective": "Review the current execution result, decide whether the incident can close, whether another execution pass is needed, or whether the work should hand off."
      }
    ],
    "transitions": [
      {
        "from": "assess",
        "on": "proceed",
        "to": "execute"
      },
      {
        "from": "assess",
        "on": "handoff",
        "to": "$handoff"
      },
      {
        "from": "execute",
        "on": "complete",
        "to": "review"
      },
      {
        "from": "review",
        "on": "complete",
        "to": "$end"
      },
      {
        "from": "review",
        "on": "needs-more-work",
        "to": "execute"
      },
      {
        "from": "review",
        "on": "handoff",
        "to": "$handoff"
      }
    ],
    "error_policy": {
      "tool_failure": {
        "action": "retry",
        "max_retries": 2,
        "on_exhausted": "fail_phase"
      },
      "phase_failure": {
        "action": "handoff"
      }
    }
  },
  "readme": "README.md",
  "license": {
    "spdx": "MIT"
  }
}
Loop graph
3 phases6 transitions5 explicit outcomes
Archetype
incident-response
Entry phase
assess
Max steps
12
Terminal targets
$handoff$end
EntryCheckpointed phaseTerminal targets
proceedhandoffcompletecompleteneeds-more-workhandoffassessexecuteimplicit completereview$handoff$end
Error policy
Tool failure
Retry up to 2 time(s), then fail_phase.
Phase failure
On phase failure: handoff.