An accelerator for your L2/L3 team — not a replacement.
Progull is being built as an agentic layer that gives existing mainframe operations teams faster diagnosis, safer remediation and a compounding playbook library. Autonomy is opt-in per abend family, and every action has an audit trail your change board can review.
Batch abends still run on human reflexes.
Mainframe incident management is mostly people, pagers and PDFs. Tickets sit open while engineers extract logs, pattern-match error codes and decide what to restart.
Manual triage at 3am
Operators page subject-matter experts, hunt logs across SDSF and JES spool, and reassemble context before a single fix can begin.
Tribal knowledge bottlenecks
Recovery depends on a small set of veterans who know the JCL, the DB2 quirks and the historical workarounds.
Recovery measured in hours
Every minute of a stuck overnight batch ripples into SLAs, downstream apps and missed business cut-offs.
Three tiers. Autonomy is earned, per abend family.
No L2 or L3 team hands over the wheel on day one. Every abend family starts at Advisory and only moves forward when the evidence in the audit trail supports it.
Advisory
The agent detects the abend, reasons over it and produces an explainable recommendation. No action is taken on z/OS. This is the default tier for every new customer and every new abend family.
Approval
The agent proposes a specific remediation and waits for a named human to click execute inside ServiceNow. Every action is logged with the reviewer's identity.
Autonomous
The agent detects, decides and acts within policy — reserved for low-risk, high-confidence abend families that have graduated out of the Approval tier with a clean track record.
Central orchestrator, domain-specific agents.
A single monolithic 'mainframe AI' doesn't survive contact with a real z/OS estate. Progull is built as a centralized orchestration layer that routes each incident to the right domain agent, then combines the result behind a risk-based human-in-the-loop gate.
Orchestrator
Receives events from the Detection layer, classifies scope, and routes to the right domain agent. Combines partial recommendations into a single proposed action for the approval gate.
JCL Agent
Owns job-control and step-level abend families — S0C7, SB37, U4038. Reasons over JCL, PROCs, dataset state and step return codes.
Db2 Agent
Reasons about SQL-911, deadlocks, buffer-pool contention and SQLCODE-driven abends. Scoped corpus, narrow surface.
CICS Agent
Covers region-level anomalies, transaction abends and connectivity failures across CICS regions.
Every recommendation flows through a risk-based approval gate before any z/OS action. Low-risk actions in Autonomous mode execute directly. Anything above the threshold routes to the Approval tier and waits.
A four-phase reasoning loop.
Every domain agent implements the same typed contract. Each phase hands off to the next through a signed, audit-logged transition.
Detection
Subscribes to JES spool, SYSLOG and OPERLOG events. Classifies abend codes the moment they appear and opens a ServiceNow incident with first-line context.
Reasoning
Fuses SYSUDUMP, JCL, recent change events, DB2 SQLCODE and historical resolutions into an explainable root-cause narrative.
Remediation
Selects a policy-approved playbook — resubmit step, recycle CICS region, hold downstream job — and stages it for the appropriate autonomy tier.
Closure
Validates recovery, attaches the full reasoning trail and evidence pack to the incident, and closes it in ServiceNow.
Autonomy with an off-switch on every step.
Policy-as-code
Every remediation maps to a versioned, reviewable playbook. Nothing executes outside the approved set.
Human-in-the-loop modes
Run agents in shadow, recommend, or auto-execute — per job class, per environment, per time window.
Full reasoning trace
Inputs, intermediate thoughts and chosen actions are persisted with the incident for audit and learning.
Observe → Reason → Act → Verify → Learn.
Every incident traverses the same typed loop. Each transition is logged, signed, and replayable against the original evidence pack.
- STEP 01
Observe
Subscribe to SYSLOG, JES spool and CICS events. Classify the abend within seconds.
- STEP 02
Reason
Fuse SYSUDUMP, JCL, SQLCODE and recent change context into an explainable hypothesis.
- STEP 03
Act
Select a policy-approved playbook. Execute under a named surrogate ID with pre-flight checks.
- STEP 04
Verify
Confirm RC=00, dataset shape and downstream invariants before declaring recovery.
- STEP 05
Learn
Promote new patterns into KB candidates. SMEs approve before they become playbooks.
Every remediation is versioned, reviewable code.
A Progull playbook is not a free-form LLM instruction. It is a typed manifest your change board approves once and the agent executes the same way every time.
Trigger conditions
Abend code, job class, LPAR, time window and confidence threshold required to match.
Pre-flight checks
Dataset state, downstream holds, change-freeze windows and dependent jobs verified before any action.
Action set
An ordered list of typed primitives — submit, hold, release, recycle — under a named surrogate ID.
Post-conditions
RC, row counts, dump absence and CICS region health re-checked before declaring recovery.
Failure escalation
If any check fails the playbook stops, opens a Sev-2 worknote and pages the assignment group.
name: pb-mf-014-s0c7-payroll
match:
abend: S0C7
job_class: PAYROLL
confidence_min: 0.85
preflight:
- assert dataset(PAY.OUT).extents_remaining > 2
- assert job(GLFEED01).state == WAITING
actions:
- hold job: GLFEED01
- submit job: PAYRUN02
step: STEP0040
parm: "CLEAN"
surrogate_id: PROGULL.PROD.PAY
postcheck:
- assert step.RC == 0
- assert dataset(PAY.OUT).row_count > 100000
on_failure:
escalate: assignment_group=MF-PAYROLL-SREAutonomy is a dial, not a switch.
Pick the mode per environment, per job class and per time window. Promote forward only when your operators trust the trail.
Advisory
Default tier. Agents observe and produce a full reasoning trail. Zero action taken on z/OS. Every abend family starts here.
Approval
Agents propose a specific remediation. A named human approves the action inside ServiceNow before it executes.
Autonomous
Reserved for low-risk families that have graduated out of Approval with a clean track record. Agents detect, decide and act within policy.
What happens when the agent is wrong.
Failure handling is designed in. The agent never silently retries; every escalation lands in ServiceNow with the same trail your auditor consumes.
Low-confidence diagnosis
Below the configured threshold, the agent halts at Recommend and pages the assignment group with the evidence pack — never executes.
Pre-flight check fails
The playbook stops on the first failed assertion, records which assertion failed and opens a Sev-2 worknote for human triage.
Post-condition fails
If RC, row counts or region health do not return as expected, the agent does not declare recovery and re-opens the incident with full diff.
Policy not matched
If no approved playbook matches, the agent writes a recommendation only. New playbooks always require a human change request.
Repeated abend pattern
After N occurrences of the same abend in a window, the agent stops auto-resolving and escalates for SME review — the playbook is no longer the answer.
Operator override
Operators can revoke an in-flight action via a single click in ServiceNow. The agent will not retry the same action without a fresh approval.