detect-dependency-risk

name	detect-dependency-risk
version	1.0.0
description	Detects cross-team and external dependency risks by analyzing issue links. Classifies risk by blocking ticket status, days to sprint end, and story points at stake. Surfaces risks before they cascade.
category	sprint-operations
trigger	Sprint health check, morning scan, mid-sprint checkpoint, dependency review, PI planning
autonomy	supervised
portability	universal
complexity	intermediate
type	detection
inputs	[{"name":"sprint_tickets","type":"structured-text","required":true,"description":"List of tickets in the active sprint. Each ticket must include: key, summary, status, assignee, priority, story_points, linked_issues. linked_issues should include: type (\"blocks\", \"is blocked by\"), key, status, and optionally team (owning team name).\n"},{"name":"sprint_context","type":"structured-text","required":true,"description":"Sprint metadata: name, start_date, end_date, days_remaining. Used for urgency classification.\n"},{"name":"team_name","type":"text","required":false,"description":"The name of the team whose sprint is being analyzed. Used to distinguish cross-team from intra-team dependencies. If not provided, all \"blocks\"/\"is blocked by\" links are analyzed.\n"}]
outputs	[{"name":"dependency_risks","type":"structured-text","description":"Prioritized list of dependency risks with blocking ticket, blocked ticket, teams involved, risk level, and recommended action.\n"},{"name":"summary","type":"text","description":"One-paragraph summary: count of at-risk dependencies, total SP exposed, highest-risk items.\n"}]
model_compatibility	["claude","gpt-4","gemini","llama-3"]

Detect Dependency Risk

Analyzes issue links in sprint tickets to identify dependencies that could block delivery. Focuses on cross-team and external dependencies where the blocking ticket is not yet complete. Produces a risk-ranked dependency table with recommended actions.

When to Use

• Daily sprint scan (to catch new dependency risks early)
• Mid-sprint checkpoints (to assess delivery risk from dependencies)
• Sprint planning (to identify dependency risks before committing)
• Cross-team synchronization meetings
• PI-level dependency reviews

What Is a Dependency Risk?

A dependency risk exists when a sprint ticket depends on another ticket (via blocks or is blocked by links) and the blocking ticket is not yet in a terminal status (Done, Closed). The risk level depends on:

1. Blocking ticket status: Is it not started, in progress, or itself blocked?
2. Days remaining in sprint: How much time is left for the dependency to resolve?
3. Story points at stake: How much work is blocked downstream?
4. Cross-team vs intra-team: Cross-team dependencies carry higher risk due to coordination overhead.

Method

Step 1: Extract all dependency links

Scan all linked_issues across sprint tickets. Collect links where:

• type is "blocks" or "is blocked by"

For each link, record:

• Blocking ticket: the ticket that must complete first
• Blocked ticket: the ticket waiting on the blocker
• Blocker status: current status of the blocking ticket
• Blocker team: team owning the blocking ticket (from team field, or inferred from ticket key prefix)

Normalize direction: if ticket A blocks ticket B, then A is the blocker and B is the blocked ticket. If ticket C is blocked by ticket D, then D is the blocker and C is the blocked ticket.

Step 2: Filter to active risks

Exclude dependencies where:

• The blocking ticket is already Done or Closed (resolved dependency)
• Both tickets are in the same sprint and assigned to the same person (personal sequencing, not a team risk)

Step 3: Classify dependency type

Classification	Criteria
External	Blocking ticket belongs to a different team (different team field or different key prefix)
Intra-team	Both tickets belong to the same team
Cascading	The blocking ticket is itself blocked by another ticket (dependency chain)

Step 4: Score risk level

For each active dependency, compute risk based on:

Factor	Low (1)	Medium (2)	High (3)
Blocker status	In Progress / In Review	To Do	Blocked
Days remaining	> 5 days	3-5 days	< 3 days
SP at stake	< 5 SP	5-8 SP	> 8 SP
Dependency type	Intra-team	External	Cascading

Risk score = sum of the four factor scores (4-12 range).

Risk Score	Risk Level	RAG
4-5	Low	Green
6-7	Medium	Amber
8-9	High	Red
10-12	Critical	Red (escalate)

Step 5: Determine recommended action

Risk Level	Recommended Action
Low	Monitor — dependency is progressing normally
Medium	Check in with blocker owner — confirm ETA and communicate to blocked team
High	Escalate to both team leads — establish daily sync on the blocking item
Critical	Escalate to management — prepare contingency plan (descope, workaround, or re-sequence)

For cross-team dependencies, always recommend explicit communication regardless of risk level.

Step 6: Detect dependency chains

If a blocking ticket is itself blocked by another ticket, flag the chain:

• Example: PLAT-100 blocks HRZ-200, but PLAT-100 is blocked by INFRA-50
• Chain: INFRA-50 → PLAT-100 → HRZ-200
• Risk level: automatically High or Critical (cascading factor = 3)

Step 7: Compile summary

Count total dependency risks by level. Sum story points exposed. Identify the highest-risk dependency for the executive summary.

Output Format

## Dependency Risks: {count} active dependencies

**Summary**: {count} dependency risks found, {critical_count} critical, {high_count} high.
{total_sp} SP at risk from unresolved dependencies with {days_remaining} days remaining.

---

### DEP-{N}: [{blocked_key}] blocked by [{blocker_key}]

- **Blocked ticket**: {blocked_key} — {blocked_summary} ({blocked_sp} SP, {blocked_status})
- **Blocking ticket**: {blocker_key} — {blocker_summary} ({blocker_status})
- **Blocker team**: {team_name or "same team"}
- **Dependency type**: {External | Intra-team | Cascading}
- **Risk level**: {Low | Medium | High | Critical} (score: {N}/12)

**Risk factors**:
- Blocker status: {status} ({score}/3)
- Days remaining: {days} ({score}/3)
- SP at stake: {sp} ({score}/3)
- Type: {type} ({score}/3)

{If cascading: "**Chain**: {chain description}"}

**Recommended Action**: {action from Step 5}

---

### Dependency Risk Summary

| Risk Level | Count | SP Exposed |
|-----------|-------|-----------|
| Critical | {N} | {SP} |
| High | {N} | {SP} |
| Medium | {N} | {SP} |
| Low | {N} | {SP} |

**Confidence**: {High | Medium | Low} — {justification}

Error Handling

• No linked_issues on any ticket: Report "No dependency links found in sprint data. Dependency risk detection requires linked_issues with blocks / is blocked by link types." Do not fabricate dependencies.
• No blocking dependencies (all resolved or no links): Report "No active dependency risks — all linked blockers are resolved." This is a positive outcome.
• Missing team data: Cannot classify as cross-team vs intra-team. Treat all as "unknown team" and note: "Team ownership data unavailable — all dependencies treated as potential cross-team risks."
• Blocker ticket not in sprint: This is normal for cross-team dependencies. Use the blocker's status from the link data. Note if the status may be stale: "Blocker status from link data — verify current status with owning team."
• Circular dependencies: If A blocks B and B blocks A, flag as: "Circular dependency detected between {A} and {B} — requires manual resolution." Classify as Critical.