Key outcomes
- Define the agent user journey and outcome
- Clarify guardrails, risks, and human touchpoints
- Prioritize metrics for success and safety
Deliverables
- Agent opportunity canvas
- Guardrail hypothesis
- Success metric stack
Prerequisites
- None
Evaluation signals
- AGENT-IDEA-001
- SAFE-GUARD-BASE
Persona fit
Assistant orchestration
View assistant playbookScout
AgentResearch horizons, regulation updates, and pattern watchlists.
- Refresh critical intel within 24 hours of change
- Maintain 95% citation accuracy
- Flag module freshness risks automatically
Coach
AgentPair with learners during modules, labs, and retrospectives.
- Median response under 2 seconds
- Satisfaction above 4.6/5
- Escalate risky experiments within 10 minutes
Critic
AgentGuardrails, evaluations, and red-team simulations.
- Detect 98% evaluation anomalies
- Zero unlogged high-severity incidents
- Attach control evidence to every flagged issue
Archivist
AgentEvidence locker, credential manifests, and knowledge graph links.
- Tag 100% deliverables with owners and signals
- Keep schema drift under 1%
- Generate credential payloads automatically
Companion
AgentHealth, pacing, and personalised nudges across squads.
- On-time nudges for 90% milestones
- Keep burnout false positives below 5%
- Publish weekly sponsor-ready progress snapshots
Navigator
AgentCTA instrumentation, sponsor digest composition, and mastery guardrails.
- Cover 95% persona CTAs every sprint
- Generate sponsor digest drafts within 5 minutes of module completion
- Hold mastery drift within one tier
Micro lessons
Signal Alignment Canvas
25 minObjective: Synthesize stakeholder inputs into a single mission brief with measurable success criteria.
Activities
- Review stakeholder interview notes
- Highlight desired outcomes vs. guardrail concerns
- Draft the mission statement
Knowledge checks
- List the top three mission outcomes.
- Which guardrail has the highest risk exposure?
Guardrail Mapping Sprint
30 minObjective: Translate organizational policies into agent-ready guardrail requirements.
Activities
- Identify policy sources
- Map policies to agent behaviours
- Capture human-in-the-loop checkpoints
Knowledge checks
- Which policy governs data retention?
- Where is human approval enforced?
Outcome Prioritisation Lab
35 minObjective: Rank outcomes and risks to guide prototype scope and evaluation focus.
Activities
- Score outcomes by impact
- Assess feasibility vs. risk
- Select success metrics
Knowledge checks
- Which metric signals guardrail drift earliest?
- What is your critical success KPI?
Knowledge points
Guardrail Taxonomy
Break risks into safety, compliance, reputational, and operational guardrails so agents can route mitigation paths.
Stakeholder Persona Matrix
Map sponsor, operator, compliance, and end-user personas to pain points and expectations to calibrate the mission brief.
Outcome Metric Stack
Select leading, lagging, and guardrail metrics so the assistant can track mission health automatically.
Micro paths featuring this module
Stand up a dependable coding agent that ships quality pull requests.
Credential alignment
This module contributes evidence across multiple credentials. See the credential framework for details.