- Home
- Skills
- Agents & Orchestration
- atlassian analyst
atlassian analyst
Systematic requirements analysis using Jira, Confluence, and PostgreSQL evidence to generate source-traceable task docs.
Free
Works with the AI Agents you already use
atlassian analyst
Example session with this skill installed
Analyze Jira issue PROJ-123. Look for related Confluence docs and check the 'orders' table in the PostgreSQL database for schema alignment.
- Read your context and instructions
- Compiled the atlassian analyst
I have completed the systems analysis for PROJ-123. I retrieved 1 parent task, 3 subtasks, 2 linked Confluence pages, and the schema for the orders table. The consolidated report is available at ~/analyst-tasks/PROJ-123/task.md.
Analysis Summary for PROJ-123
The requirement involves modifying the order status workflow. There is a discrepancy between the Confluence documentation (stating 5 statuses) and the PostgreSQL order_status enum (showing 7 statuses).
Data Schema (PostgreSQL)
| Table | Column | Type | Constraints |
|---|---|---|---|
| orders | id | uuid | PRIMARY KEY |
| orders | status | varchar(20) | NOT NULL |
| orders | updated_at | timestamp | DEFAULT now() |
Requirement Traceability
| ID | Requirement | Source | Verification Criteria |
|---|---|---|---|
| REQ-01 | Status transition must be logged | J1, C2 | Check order_history table |
| REQ-02 | Prevent rollback to 'Pending' | J1 | JQL: status != 'Pending' |
PlantUML Process Flow
@startuml
participant "Client" as C
participant "OrderService" as S
database "PostgreSQL" as DB
C -> S: updateStatus(orderId, newStatus)
S -> DB: SELECT status FROM orders WHERE id = orderId
DB -> S: current_status
alt valid transition
S -> DB: UPDATE orders SET status = newStatus
S -> C: 200 OK
else invalid
S -> C: 400 Bad Request
end
@enduml
Next steps
- Review the "Contradictions" section in
task.mdregarding the 2 extra status codes found in the database. - Verify if the
order_historytable needs a new index for the proposed reporting requirements. - Use the provided PlantUML code to update the project's architecture diagrams.
Connects securely to your tools. The creator never sees your data.
About this skill
Manual systems analysis across fragmented documentation is a primary cause of technical debt and misaligned requirements. This skill automates the discovery and synthesis of project context by auditing Jira issues, Confluence pages, and PostgreSQL schemas simultaneously. It eliminates guesswork by cross-referencing task descriptions with live database structures and documented processes, producing a single, source-traceable technical requirement document.
What it does
- Multi-source extraction reads complete issue hierarchies from Jira, full-text documentation from Confluence, and live schemas from PostgreSQL.
- Traceable synthesis maps every requirement, business rule, and data model back to specific source IDs (J1, C1, P1).
- Conflict detection flags discrepancies between Jira tasks, Confluence docs, and actual database states for manual review.
- Diagram generation produces PlantUML sequence and ER diagrams based on observed system interactions and table relationships.
- Full audit logging records exact search criteria, row counts, and pagination status to ensure data completeness.
How it works
- Scope definition starts with a Jira key or JQL filter to establish the initial boundary of the analysis.
- Concurrent retrieval executes MCP calls to Jira for tickets, Confluence for wikis, and PostgreSQL for schema inspection and data sampling.
- Requirement mapping builds a traceability matrix linking business rules to their origins in the three sources.
- Markdown assembly generates a
task.mdfile containing the executive summary, technical specs, PlantUML blocks, and a raw source appendix.
Frameworks & tools
This skill requires MCP servers for Jira, Confluence, and PostgreSQL. It generates standard Markdown and PlantUML syntax.
Why this beats prompting it yourself
Standard prompts often hallucinate missing context or fail to handle the pagination required for deep Jira histories and large DB schemas. This skill enforces a strict data-first approach, requiring evidence from all three sources before finalizing requirements and explicitly logging what it could not find.
Use cases
- Feature discovery for new developers joining a project with undocumented legacy behavior.
- Pre-sprint grooming to verify that Jira descriptions match the current database reality.
- Technical debt auditing to identify where code/data structures have diverged from Confluence specifications.
- Automated documentation generating a comprehensive technical baseline for complex bug fixes.
Known limitations
PostgreSQL access is strictly read-only and requires a pre-configured MCP connection. Large attachments or complex Confluence macros that cannot be converted to text will be logged as links only.
Requires active MCP servers for Jira, Confluence, and PostgreSQL. PostgreSQL is read-only. Does not render PlantUML; provides code blocks only. Output path defaults to ~/analyst-tasks/.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 4 days ago
- Free with an Agensi account
Requires
Permissions
Allowed Hosts