Product Owner Skill

A Claude-native product owner that provides daily standups, priority recommendations, and cross-project awareness without artificial sprint boundaries. Enhanced with AI-powered analysis, metrics tracking, and intelligent dependency detection.

Contextd Integration (Optional)

If contextd MCP is available:

• memory_record for standup patterns, blockers, and prioritization decisions
• Cross-session priority tracking with decision accuracy measurement
• memory_search to find past standup context and recurring patterns
• remediation_search for known blockers and their resolutions

If contextd is NOT available:

• GitHub queries work normally
• No cross-session memory (stateless standups)
• No decision tracking or pattern learning

Philosophy

• Recommend, don't dictate: Present priorities; user decides what to act on
• Continuous flow: No sprints or ceremonies - priorities adjust daily
• Cross-project awareness: Dependencies across repos get flagged and prioritized
• Memory persistence: Use contextd if available (optional)
• Data-driven decisions: Leverage DORA metrics and WSJF for objective prioritization
• Learn from history: Track decision accuracy to improve recommendations over time

When to Use

Trigger	Use Case
/standup	Daily standup report with AI-generated blockers
/standup --metrics	Include DORA metrics and velocity
/standup --platform	Cross-project view with dependency map
"what should I work on?"	WSJF-ranked priority recommendations
"project status"	Current state with risk scoring
"what's blocking?"	AI-powered blocker analysis
"show velocity"	Velocity tracking and forecasting
"show metrics"	DORA metrics dashboard

Data Sources

GitHub (via MCP)

Query	Purpose
list_pull_requests	Open PRs, review states, ages
list_issues	Priority-labeled issues
list_commits	Recent activity on main
list_branches	Detect stale branches
get_commit	Deployment tracking for DORA
search_issues	Cross-repo dependency detection

GitHub Projects v2 (via GraphQL)

Query	Purpose
Project items	Sprint/iteration tracking
Custom fields	Story points, priority, status
Project views	Board and table layouts

GraphQL Example:

query {
  organization(login: "fyrsmithlabs") {
    projectV2(number: 1) {
      items(first: 100) {
        nodes {
          content { ... on Issue { title number } }
          fieldValues(first: 10) {
            nodes { ... on ProjectV2ItemFieldNumberValue { number } }
          }
        }
      }
    }
  }
}

contextd (Cross-Session Memory)

Query	Purpose
checkpoint_list/resume	Yesterday's state
memory_search	Recurring patterns, blockers
remediation_search	Known issues and fixes
memory_record	Store prioritization decisions

---

AI-Powered Analysis

Risk Scoring

Calculate risk score (0-100) based on technical complexity and business impact:

Risk Score = (Technical Complexity * 0.4) + (Business Impact * 0.4) + (Time Sensitivity * 0.2)

Technical Complexity factors:
- Lines changed: >500 = HIGH, >100 = MEDIUM, else LOW
- Files touched: >10 = HIGH, >3 = MEDIUM, else LOW
- Cross-service changes: +20 points
- Database migrations: +25 points
- Security-sensitive areas: +30 points

Business Impact factors:
- Label priority:critical = 40, priority:high = 30, else 10
- Customer-facing: +20 points
- Revenue-impacting: +25 points
- Compliance-related: +30 points

Time Sensitivity factors:
- Days until deadline (if any)
- External dependencies timing
- Release train alignment

Risk Categories:

Score	Category	Action
80-100	CRITICAL	Immediate attention, may need escalation
60-79	HIGH	Prioritize this session
40-59	MEDIUM	Schedule within week
0-39	LOW	Normal queue

Semantic Dependency Detection

AI analyzes issue/PR content for implicit dependencies beyond explicit references:

Patterns detected:

• Shared code paths (same files modified)
• Common data models or schemas
• API contract dependencies
• Feature flag interdependencies
• Deployment order requirements

Example detection:

Issue #45: "Add user avatar upload"
Issue #52: "Implement image resizing service"
  -> DETECTED: #45 likely depends on #52 (image processing capability)

Automated Blocker Identification

Scan issue comments and PR reviews for blocker signals:

Signal patterns:

# Explicit blockers
blocked by|waiting on|depends on|can't proceed

# Implicit blockers (in comments)
need.*first|requires.*before|after.*merged
stuck on|no progress|help needed

# Review blockers
changes requested.*\d+ days ago
approved but.*merge conflict
CI.*failing.*\d+ hours

AI Comment Analysis:

PR #42 Comments Analysis:
  - @alice (2 days ago): "This needs the auth refactor first"
  - @bob (1 day ago): "Still waiting on API spec"
  -> BLOCKERS DETECTED: Auth refactor, API spec finalization

---

WSJF Prioritization

Weighted Shortest Job First for objective prioritization:

WSJF Score = Cost of Delay / Job Size

Cost of Delay = User Value + Time Criticality + Risk Reduction

Where:
- User Value: Business impact (1-10)
- Time Criticality: Urgency decay (1-10)
- Risk Reduction: Technical/business risk mitigated (1-10)
- Job Size: Estimated effort in story points or T-shirt sizes

WSJF Calculation Example

Issue: Implement OAuth2 login
  User Value: 8 (high user demand)
  Time Criticality: 6 (competitor launching similar)
  Risk Reduction: 7 (security improvement)
  Job Size: 5 (medium complexity)
  WSJF = (8 + 6 + 7) / 5 = 4.2

Issue: Fix typo in docs
  User Value: 2 (minor inconvenience)
  Time Criticality: 1 (no deadline)
  Risk Reduction: 1 (no risk)
  Job Size: 1 (trivial)
  WSJF = (2 + 1 + 1) / 1 = 4.0

Result: OAuth2 login prioritized higher despite larger size

Auto-WSJF from Labels

Map GitHub labels to WSJF components:

Label	Component	Value
priority:critical	Time Criticality	10
priority:high	Time Criticality	7
security	Risk Reduction	9
customer-facing	User Value	8
tech-debt	Risk Reduction	5
size:XL	Job Size	13
size:L	Job Size	8
size:M	Job Size	5
size:S	Job Size	3
size:XS	Job Size	1

---

DORA Metrics Integration

Track and display key DevOps Research and Assessment metrics:

Metrics Definitions

Metric	Definition	Calculation
Deployment Frequency	How often code deploys to production	Deploys per day/week
Lead Time for Changes	Time from commit to production	PR open -> merge -> deploy
Mean Time to Recovery	Time to restore service after incident	Incident open -> resolved
Change Failure Rate	% of deployments causing failures	Failed deploys / total deploys

Data Collection

Deployment Frequency:

# Count merges to main/production in time period
mcp__MCP_DOCKER__list_commits(
  owner: "<org>",
  repo: "<repo>",
  sha: "main",
  since: "<period_start>"
)

Lead Time for Changes:

# For each merged PR:
Lead Time = PR merged_at - first_commit_at + deploy_delay

Change Failure Rate:

# Track labels/issues indicating failures
mcp__MCP_DOCKER__list_issues(
  labels: "incident,hotfix,rollback"
)

Metrics Display Format

+-------------------------------------------------------------+
| DORA Metrics: <repo> (Last 30 days)                         |
+-------------------------------------------------------------+
| Deployment Frequency:  2.3/day  ########-- Elite            |
| Lead Time for Changes: 4.2 hrs  #######--- High             |
| Mean Time to Recovery: 1.1 hrs  ########-- Elite            |
| Change Failure Rate:   8%       #######--- High             |
+-------------------------------------------------------------+
| Trend: ^ Improving (vs last 30 days)                        |
| Team Performance: HIGH PERFORMER                            |
+-------------------------------------------------------------+

Performance Benchmarks

Metric	Elite	High	Medium	Low
Deploy Frequency	Multiple/day	Daily-Weekly	Weekly-Monthly	Monthly+
Lead Time	<1 hour	<1 day	<1 week	>1 week
MTTR	<1 hour	<1 day	<1 week	>1 week
Change Failure Rate	<5%	5-10%	10-15%	>15%

---

Velocity Tracking

Velocity Calculation

Velocity = Story Points Completed / Time Period

Rolling Average (recommended):
  velocity_avg = (v_current * 0.4) + (v_prev1 * 0.3) + (v_prev2 * 0.2) + (v_prev3 * 0.1)

Velocity Sources

From GitHub:

• Closed issues with size labels
• Merged PRs with story points
• GitHub Projects v2 custom fields

From contextd:

• Historical velocity records
• Pattern trends over time

Forecasting

Remaining Work = Sum of story points in backlog
Predicted Completion = Remaining Work / Velocity Average

Example:
  Backlog: 45 story points
  Velocity: 15 points/week
  Forecast: 3 weeks to completion

  With 80% confidence interval:
    Best case: 2.4 weeks (velocity + 25%)
    Worst case: 4.5 weeks (velocity - 33%)

Velocity Display

Velocity Trend (Last 4 Weeks):
  Week -3: ########---- 12 pts
  Week -2: ##########-- 15 pts
  Week -1: #######----- 11 pts
  Current: #########--- 14 pts

  Average: 13 pts/week
  Trend: Stable (+/-15%)

  Backlog: 52 pts remaining
  Forecast: 4 weeks (3-5 week range)

---

Cross-Project Intelligence

Dependency Mapping

Automatically map dependencies across repositories:

Dependency Graph: fyrsmithlabs
+------------------+     +------------------+
|   marketplace    |---->|    contextd      |
|  (v1.6.0)        |     |  (v1.0.0)        |
+------------------+     +------------------+
         |                       |
         v                       v
+------------------+     +------------------+
|   fs-design      |     |   mcp-server     |
|  (v1.0.0)        |     |  (v0.9.0)        |
+------------------+     +------------------+

Legend:
  ---> Runtime dependency
  ...> Dev dependency
  !!!> Blocking dependency (needs action)

Shared Component Impact Analysis

Detect when changes affect multiple projects:

Detection triggers:

• Shared library version changes
• API contract modifications
• Database schema changes
• Configuration format changes

Impact report:

SHARED COMPONENT ALERT: contextd MCP types v1.3.0
  Affected repositories:
    - marketplace (uses v1.2.0) - UPDATE NEEDED
    - claude-code-plugin (uses v1.3.0) - Current
    - internal-tools (uses v1.1.0) - UPDATE NEEDED

  Breaking changes in v1.3.0:
    - memory_record signature changed
    - New required field: project_id

Team Capacity Allocation

Track and recommend work distribution:

Team Capacity Overview:
  Total: 5 contributors active this week

  Allocation:
    marketplace:    3 contributors (60%)
    contextd:       2 contributors (40%)
    fs-design:      1 contributor  (20%)
                    ^ overlap detected

  Recommendations:
    - contextd has 2 HIGH priority items but limited capacity
    - Consider cross-training for fs-design (single point of failure)

---

Enhanced Standup Features

AI-Generated Blockers from Comments

Automatically extract blockers from issue/PR discussions:

BLOCKER ANALYSIS (AI-Generated):

PR #42 - Add OAuth2 login
  Explicit: None stated
  Detected from comments:
    - "Waiting on security review" (@alice, 2 days ago)
    - "Need API docs updated first" (@bob, 1 day ago)
  Confidence: HIGH
  Suggested action: Follow up with security team

Issue #38 - Performance regression
  Explicit: "blocked by #35"
  Detected from comments:
    - "Can't reproduce without staging access" (@carol, 3 days ago)
  Confidence: MEDIUM
  Suggested action: Provision staging access for @carol

Automated PR Status Summary

Rich PR status with actionable insights:

PR STATUS SUMMARY:

Ready to Merge (2):
  #42 OAuth2 login - Approved 2 days ago, CI passing
       ^ STALE: Consider merging or comment on delay
  #45 Fix typo - Approved today, CI passing

In Review (3):
  #48 Add metrics dashboard - 1 approval, waiting on @alice
       Last activity: 4 hours ago
  #50 Refactor auth module - Changes requested by @bob
       ^ ACTION NEEDED: Address review comments
  #51 Update deps - No reviews yet (opened 1 hour ago)

Draft (1):
  #52 WIP: New feature - Draft, 15 commits
       Last push: 30 minutes ago

Sprint Burndown Context

When using iterations/sprints:

Sprint Progress (if applicable):
+-----------------------------------------------+
| Sprint: 2026-W05 (Jan 27 - Feb 2)             |
+-----------------------------------------------+
| Day 2 of 5                                    |
|                                               |
| Planned: 34 pts  | Completed: 8 pts           |
| Remaining: 26 pts                             |
|                                               |
| Burndown:                                     |
|   Ideal:  ##########################          |
|   Actual: ####----------------------          |
|                                               |
| Status: BEHIND (ideal: 13.6 pts)              |
| Risk: HIGH - May not complete sprint goal     |
|                                               |
| Recommendations:                              |
|   - Reduce scope: Consider moving #48 to next |
|   - Focus: Complete #42 today (ready to merge)|
+-----------------------------------------------+

---

External Integrations

Slack/Teams Notification Patterns

Send standup summaries to team channels:

Slack Webhook Format:

{
  "channel": "#dev-standup",
  "username": "Product Owner Bot",
  "icon_emoji": ":clipboard:",
  "blocks": [
    {
      "type": "header",
      "text": {
        "type": "plain_text",
        "text": "Daily Standup: marketplace"
      }
    },
    {
      "type": "section",
      "fields": [
        {"type": "mrkdwn", "text": "*CRITICAL:* 0"},
        {"type": "mrkdwn", "text": "*HIGH:* 2"},
        {"type": "mrkdwn", "text": "*PRs Ready:* 2"},
        {"type": "mrkdwn", "text": "*Blockers:* 1"}
      ]
    },
    {
      "type": "section",
      "text": {
        "type": "mrkdwn",
        "text": "*Top Recommendation:* Merge PR #42 (approved, waiting 2 days)"
      }
    }
  ]
}

Teams Adaptive Card:

{
  "type": "AdaptiveCard",
  "body": [
    {
      "type": "TextBlock",
      "size": "Large",
      "weight": "Bolder",
      "text": "Daily Standup: marketplace"
    },
    {
      "type": "ColumnSet",
      "columns": [
        {"type": "Column", "items": [{"type": "TextBlock", "text": "CRITICAL: 0"}]},
        {"type": "Column", "items": [{"type": "TextBlock", "text": "HIGH: 2"}]}
      ]
    }
  ]
}

Jira Bridge for Hybrid Teams

Map between GitHub Issues and Jira:

Jira Field Mapping:

GitHub	Jira
priority:critical	Priority: Highest
priority:high	Priority: High
size:XL	Story Points: 13
Issue labels	Jira labels
Milestone	Fix Version

Sync Pattern:

# Query Jira for linked issues
jira_issues = jira.search("project = PROJ AND 'GitHub Issue' is not EMPTY")

# Correlate with GitHub
for issue in jira_issues:
  gh_ref = issue.customfield_github
  gh_issue = github.get_issue(gh_ref)

  # Sync status
  if gh_issue.state == "closed" and issue.status != "Done":
    notify("Jira issue out of sync with GitHub")

Hybrid Standup Report:

Cross-Platform Status:
  GitHub Issues: 12 open (3 HIGH)
  Jira Tickets: 8 open (2 HIGH)
  Synced: 5 issues linked

  Discrepancies:
    - PROJ-123 (Jira: In Progress) != #45 (GitHub: Closed)
      ^ ACTION: Update Jira status

---

contextd Decision Tracking

Recording Prioritization Decisions

Track why items were prioritized:

mcp__contextd__memory_record(
  project_id: "fyrsmithlabs/marketplace",
  title: "Prioritization decision: OAuth2 over metrics",
  content: |
    Context: Both OAuth2 (#42) and metrics (#48) were HIGH priority
    Decision: Prioritized OAuth2 first
    Reasoning:
      - WSJF: OAuth2 (4.2) > metrics (3.8)
      - Security benefit from OAuth2
      - External deadline for OAuth2 (partner launch)
    Expected outcome: OAuth2 merged by EOD
  outcome: "pending",
  tags: ["standup", "prioritization", "decision"]
)

Tracking Decision Accuracy

Measure prioritization quality over time:

mcp__contextd__memory_outcome(
  memory_id: "<decision_memory_id>",
  outcome: "success",
  notes: "OAuth2 merged on time, partner launch successful"
)

Accuracy Report:

Prioritization Accuracy (Last 30 days):
  Total decisions tracked: 24
  Correct predictions: 19 (79%)

  Patterns:
    - Security items: 95% accuracy (prioritize well)
    - Time estimates: 65% accuracy (often underestimate)
    - Blocker detection: 72% accuracy (improve detection)

  Recommendations:
    - Add 20% buffer to time estimates
    - More aggressive blocker follow-up

Searching Past Standup Context

Find relevant historical context:

mcp__contextd__memory_search(
  query: "OAuth implementation blockers",
  tags: ["standup", "blocker"],
  limit: 5
)

Results:
  1. "Blocker: OAuth2 waiting on security review" (2 days ago)
     Resolution: Escalated to security lead, resolved in 4 hours

  2. "Blocker: OAuth2 library compatibility" (1 week ago)
     Resolution: Downgraded to v2.1.0, documented in ADR

---

Priority Classification

CRITICAL (Immediate Attention)

Items requiring immediate attention. Stop current work if needed.

Criteria:

• Security vulnerabilities (label: security, vulnerability)
• Failing CI on main/production branches
• PRs blocked >24h with no activity
• Production incidents or outages
• Data integrity issues
• Risk score >= 80

Action: Address before anything else.

HIGH (Today's Focus)

Primary work items for the current session.

Criteria:

• PRs approved and ready to merge
• Issues with priority:high or priority:critical label
• Items explicitly marked for today in last checkpoint
• Carried over items from previous standup
• Time-sensitive deliverables
• WSJF score in top quartile
• Risk score 60-79

Action: Complete these during current session.

DEPENDENCY ALERT (Cross-Project)

Items with cross-repository dependencies.

Detection patterns in issue/PR bodies:

- "blocked by <repo>#<number>"
- "depends on <repo>"
- "waiting for <repo>"
- "upstream: <repo>"
- References to other fyrsmithlabs repos

Semantic detection:

• Shared file modifications across repos
• API contract consumers
• Shared library version conflicts

Action: Flag prominently, include in cross-project view.

MEDIUM (This Week)

Important but not urgent items.

Criteria:

• PRs in active review
• Issues with priority:medium or no priority label
• Planned work mentioned in checkpoints
• Technical debt items
• Risk score 40-59

Action: Work on after HIGH items complete.

CARRIED OVER (Tracking)

Items from previous sessions not yet completed.

Detection:

• Compare current checkpoint to previous
• Items mentioned but not closed
• Stale branches (>24h without PR)

Action: Review whether still relevant, re-prioritize or close.

Standup Report Format

Standard Report

+-------------------------------------------------------------+
| Daily Standup: <owner>/<repo>                               |
| <date> at <time>                                            |
+-------------------------------------------------------------+

Yesterday:
  - <accomplishments from last checkpoint>
  - <velocity: n issues closed, m PRs merged>

CRITICAL (<count>):
  * <item> - <status/reason> [Risk: <score>]

HIGH (<count>):
  * <item> - <status> [WSJF: <score>]

DEPENDENCY ALERT:
  ! <description of cross-project dependency>

MEDIUM (<count>):
  * <item> - <status>

CARRIED OVER from yesterday:
  * <item> - <reason still open>

AI-Detected Blockers:
  * <blocker from comment analysis>

Recommendations:
  1. <actionable next step>
  2. <actionable next step>
  3. <actionable next step>

-------------------------------------------------------------

Metrics Report (--metrics flag)

+-------------------------------------------------------------+
| Daily Standup: <owner>/<repo>                               |
| <date> at <time>                                            |
+-------------------------------------------------------------+

[Standard report sections...]

DORA Metrics (30-day rolling):
  Deploy Frequency:   2.3/day  [Elite]
  Lead Time:          4.2 hrs  [High]
  MTTR:               1.1 hrs  [Elite]
  Change Failure:     8%       [High]

  Trend: Improving (+12% vs last period)

Velocity:
  This week: 14 pts (target: 15)
  Rolling avg: 13 pts/week
  Backlog: 52 pts | Forecast: 4 weeks

-------------------------------------------------------------

Brief Report (for scripting)

[<repo>] CRIT:<n> HIGH:<m> MED:<k> | PRs:<open>,<ready> | Blocked: <description> | DORA: <performance_tier>

Platform Report (cross-project)

+-------------------------------------------------------------+
| Platform Standup: fyrsmithlabs                              |
| <date> at <time>                                            |
+-------------------------------------------------------------+

Cross-Project Dependencies:
  ! <repo> -> <repo>: <description>
  OK <repo> -> <repo>: No blockers

Dependency Graph:
  marketplace --> contextd --> mcp-server
       |
       v
  fs-design

Shared Component Alerts:
  * contextd types v1.3.0 needs update in marketplace

Project Summaries:
  [<repo>] CRIT:<n> HIGH:<m> MED:<k> | Focus: <summary>

Team Capacity:
  marketplace: 3 active | contextd: 2 active

Recommended Focus Order:
  1. <repo> - <reason>
  2. <repo> - <reason>

-------------------------------------------------------------

Cross-Project Discovery

Finding fyrsmithlabs Repos

Priority order for discovering repos:

1. Config file (if exists):

   ~/.fyrsmithlabs/repos.json
   {
     "repos": [
       "fyrsmithlabs/contextd",
       "fyrsmithlabs/marketplace"
     ]
   }
   

2. Local directory scan:

   ls ~/projects/fyrsmithlabs/
   

3. GitHub org query:

   mcp__MCP_DOCKER__search_repositories(
     query: "org:fyrsmithlabs"
   )
   

Dependency Detection

Scan issue/PR content for patterns:

# Cross-repo references
fyrsmithlabs/\w+#\d+
blocked by .+#\d+
depends on .+
waiting for .+
upstream: .+

contextd Integration

Checkpoint Schema

Store standup state for next session:

checkpoint:
  name: "standup-YYYY-MM-DD"
  summary: "CRIT:0 HIGH:2 MED:5. Focus: Complete auth PR"
  context: |
    PRs: #42 (auth, ready), #38 (hooks, in review)
    Issues: #15 (high), #20 (medium), #22 (medium)
    Blockers: contextd v1.3 release
    DORA: Elite (2.3 deploys/day)
    Velocity: 14 pts this week
  full_state: |
    <complete standup output for comparison>

Memory Recording

Record notable events:

Persistent blockers:

mcp__contextd__memory_record(
  project_id: "<project>",
  title: "Persistent blocker: <description>",
  content: "Blocked on <item> for <n> days. First seen: <date>",
  outcome: "failure",
  tags: ["standup", "blocker", "persistent"]
)

Velocity patterns:

mcp__contextd__memory_record(
  project_id: "<project>",
  title: "Velocity trend: <direction>",
  content: "Closed <n> issues this week vs <m> last week",
  outcome: "success",
  tags: ["standup", "velocity"]
)

Prioritization decisions:

mcp__contextd__memory_record(
  project_id: "<project>",
  title: "Prioritization: <item_a> over <item_b>",
  content: "WSJF scores: <a>=<score>, <b>=<score>. Chose <a> because <reason>",
  outcome: "pending",
  tags: ["standup", "prioritization", "decision"]
)

Stale Work Detection

Stale PRs

Criteria:

• Open >48h with no review activity
• Approved but not merged >24h
• Changes requested but no updates >48h

Action: Flag in CARRIED OVER, suggest follow-up.

Stale Branches

Criteria:

• Branch exists with commits not on main
• No open PR associated
• Last commit >24h ago

Action: Prompt to create PR or delete branch.

Stale Issues

Criteria:

• Assigned but no activity >7 days
• In progress label but no linked PR

Action: Flag in MEDIUM, suggest status update.

Error Handling

Scenario	Handling
GitHub API unavailable	Use cached data from last checkpoint
No Git remote	Skip GitHub, show contextd-only standup
First standup (no checkpoint)	Show current state, note "first standup"
Empty repo (no issues/PRs)	Prompt "Clean slate - what's the first task?"
Rate limited	Graceful degradation with warning
DORA data insufficient	Show "Insufficient data" for affected metrics
Jira sync fails	Continue with GitHub-only data, warn user

Mandatory Checklist

Every standup MUST complete:

• Load contextd checkpoint (or note first standup)
• Query GitHub PRs and issues
• Detect cross-project dependencies
• Calculate WSJF scores for HIGH items
• Run AI blocker detection on comments
• Classify items by priority with risk scores
• Generate recommendations
• Save checkpoint for next standup
• Record blockers to memory (if any)
• Record prioritization decisions (for accuracy tracking)

Red Flags - STOP and Reconsider

Thought	Reality
"Skip checkpoint - not important"	Checkpoint enables tomorrow's comparison
"No need to check dependencies"	Cross-project blockers are often invisible
"Just show raw GitHub data"	Synthesis and prioritization is the value
"CRITICAL for everything"	Overuse dilutes urgency
"Skip contextd - just use GitHub"	Memory provides continuity across sessions
"Ignore WSJF - trust gut"	Data-driven decisions reduce bias
"Skip blocker detection"	AI catches what humans miss in comments
"DORA metrics are overhead"	Metrics drive continuous improvement

Attribution

Original research synthesized from:

• Digital Scrum Master patterns (episodic memory)
• CrewAI role-based agents (structured autonomy)
• GitHub Copilot Agent (native integration)
• fyrsmithlabs contextd (cross-session memory)
• Accelerate book (DORA metrics)
• SAFe framework (WSJF prioritization)

See CREDITS.md for full attribution.