Skip to main content
Glama

devflow-mcp

Local, read-only MCP server that securely serves Jira (Server/Data Center) data to Claude Code — without ever exposing your PAT/token to the model. (GitLab tools + the /devflow skill land in later phases; see the plan.)

Plan: plans/260625-1353-jira-ai-dev-mcp/

Why

The token lives only inside this server's HTTP client (read from an env var). Claude calls tools like jira_get_ticket_bundle("PROJ-123") and gets back trimmed JSON — it never sees the credential, and the credential is never a tool argument.

Related MCP server: jira-mcp-server

Requirements

  • Python ≥ 3.10

  • A Jira Server/DC Personal Access Token (Bearer). Profile → Personal Access Tokens.

Setup

cd devflow-mcp
python3 -m venv .venv
source .venv/bin/activate
pip install -e .            # or: pip install -e ".[dev]" for tests

cp .env.example .env        # then edit JIRA_BASE_URL + JIRA_PAT

.env (gitignored) for local runs:

JIRA_BASE_URL=https://jira.your-company.com
JIRA_PAT=xxxxxxxxxxxx

Smoke test (MCP Inspector)

fastmcp dev server.py
# open the Inspector URL, call jira_get_ticket_bundle with a real issue key

First, call jira_get_fields once and note the custom field IDs for Sprint / Epic Link / Story Points — needed in later phases.

Register in Claude Code

Add to your project .mcp.json (commit this — it has NO secrets, only ${VAR} refs):

{
  "mcpServers": {
    "devflow": {
      "command": "/ABS/PATH/devflow-mcp/.venv/bin/python",
      "args": ["/ABS/PATH/devflow-mcp/server.py"],
      "env": {
        "JIRA_BASE_URL": "${JIRA_BASE_URL}",
        "JIRA_PAT": "${JIRA_PAT}",
        "GITLAB_URL": "${GITLAB_URL}",
        "GITLAB_TOKEN": "${GITLAB_TOKEN}"
      }
    }
  }
}

Put the real values in your shell (~/.zshrc). GitLab is optional — omit it and GitLab tools self-disable:

export JIRA_BASE_URL="https://jira.your-company.com"
export JIRA_PAT="xxxxxxxxxxxx"
export GITLAB_URL="https://gitlab.your-company.com"   # self-hosted; /api/v4 appended automatically
export GITLAB_TOKEN="glpat-xxxxx"                       # scopes: read_api, read_repository

Restart Claude Code, then the tools appear.

Tools (21, all read-only)

Jira core

Tool

Purpose

jira_get_ticket_bundle(key)

primary: issue + subtasks + links + comments + attachment metadata in one call

jira_get_issue(key, fields?)

single issue, trimmed

jira_get_comments(key, limit?)

recent comments

jira_get_subtasks(key) / jira_get_linked_issues(key)

subtasks / linked issues

jira_search(jql, fields?, limit?)

JQL search (paginated, capped)

jira_get_fields()

resolve custom field IDs (setup)

Jira agile — jira_get_epic, jira_get_epic_issues, jira_get_sprint, jira_get_sprint_issues

Jira attachments (image analysis) — jira_get_attachments(key), jira_download_attachment(key, attachment_id) → downloads to a local path (token hidden) so Claude can read mockups/screenshots with vision.

GitLab (self-hosted, optional) — gitlab_search_mrs(ticket_key), gitlab_get_mr, gitlab_get_mr_commits, gitlab_get_commits, gitlab_get_commit_diff, gitlab_get_pipeline_status, gitlab_search_code, gitlab_get_file.

Security notes

  • Secrets are read from env only; never accepted as args, never returned, masked in errors.

  • Read-only: no write/transition/comment endpoints are implemented.

  • JIRA_VERIFY_SSL=false only for trusted internal self-signed certs.

Troubleshooting

  • Exits "Missing required environment variable" → set JIRA_BASE_URL / JIRA_PAT.

  • 401/403 → check PAT validity + project permissions.

  • SSL error on internal host → set JIRA_VERIFY_SSL=false (trusted hosts only).

Available Tools

22 tools
gitlab_get_commit_diffA

Get a commit's diff (read-only). Truncated for token efficiency.

Inputs: sha (required), project (required). Returns: object

gitlab_get_commitsB

List recent commits for a project branch/ref (read-only).

Inputs: project (required), ref, limit. Returns: object

gitlab_get_fileA

Read a repo file's content (read-only). Truncated for token efficiency.

Inputs: project (required), file_path (required), ref. Returns: object

gitlab_get_mrB

Get a single merge request (read-only).

Inputs: mr_iid (required), project (required). Returns: object

gitlab_get_mr_commitsB

List commits in a merge request (read-only).

Inputs: mr_iid (required), project (required), limit. Returns: object

gitlab_get_pipeline_statusA

Get latest pipeline status for a branch/ref or a merge request (read-only).

Provide either ref (branch) or mr_iid.

Inputs: project (required), ref, mr_iid. Returns: object

gitlab_search_codeA

Search code (blobs) across GitLab (read-only). Degrades gracefully.

Use this only for repos NOT checked out locally; for local repos prefer Claude's native Grep/Glob. Requires Advanced Search on the instance.

Inputs: query (required), limit, project. Returns: object

gitlab_search_mrsA

Find merge requests referencing a ticket key (read-only, heuristic).

Searches MR title + description for the key. Linkage is a HEURISTIC — results are candidates, not guaranteed matches.

Inputs: ticket_key (required), limit, project. Returns: object

jira_download_attachmentA

Download one Jira attachment to a LOCAL file and return its path (read-only).

The download uses the server's authenticated client (token stays hidden). Claude can then read the returned path with vision, or pass it to the ai-multimodal skill, to analyze an image/mockup attached to the task.

Inputs: key (required), attachment_id (required). Returns: object

jira_get_attachmentsA

List attachment metadata on a Jira issue (read-only, no download).

Use this to see whether a task has images/mockups worth analyzing. Download a specific one with jira_download_attachment.

Inputs: key (required). Returns: object

jira_get_commentsA

Fetch comments on a Jira issue (read-only). Returns the most recent limit.

Inputs: key (required), limit. Returns: object

jira_get_epicA

Get a Jira epic by key/id (read-only).

Inputs: epic_key (required). Returns: object

jira_get_epic_issuesC

List issues belonging to an epic (read-only).

Inputs: epic_key (required), limit. Returns: object

jira_get_fieldsA

List Jira field metadata to resolve custom field IDs (read-only).

Run once per instance during setup to discover IDs for Sprint, Epic Link, Story Points, etc. (these differ per Jira instance).

Inputs: no named parameters. Returns: object

jira_get_issueB

Fetch a single Jira issue (read-only).

Inputs: key (required), fields. Returns: object

jira_get_linked_issuesA

List issues linked to a Jira issue (read-only), with relation + direction.

Inputs: key (required). Returns: object

jira_get_sprintB

Get a sprint by id (read-only).

Inputs: sprint_id (required). Returns: object

jira_get_sprint_issuesB

List issues in a sprint (read-only).

Inputs: sprint_id (required), limit. Returns: object

jira_get_subtasksA

List subtasks of a Jira issue (read-only).

Inputs: key (required). Returns: object

jira_get_ticket_bundleA

One-call ticket context for analysis (read-only): the PRIMARY tool for /jira-dev.

Returns the issue + subtasks + linked issues + recent comments in a single token-efficient payload, avoiding multiple round-trips.

Inputs: key (required). Returns: object

jira_my_tasksA

List MY tasks: assigned to me, in active statuses (default: To Do, Doing).

This is the default local-dev working set — call with no key to see what to work on. Override statuses via env JIRA_ACTIVE_STATUSES.

Inputs: limit, extra_jql. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.7/5.0

    Scored across 22 tools

    Disambiguation5/5

    Each tool has a unique, well-defined purpose. GitLab and Jira tools are cleanly separated by prefixes, and within each domain, tools target distinct resources (e.g., commits, MRs, pipelines, issues, attachments) with no functional overlap.

    Naming Consistency5/5

    All tools follow a consistent `domain_verb_noun` pattern in snake_case. GitLab tools start with `gitlab_` and Jira tools with `jira_`, and verbs like `get_`, `search_`, `download_` are used uniformly.

    Tool Count4/5

    With 22 tools, the server covers two distinct domains (GitLab and Jira) comprehensively. While slightly above the typical 3-15 range, each tool addresses a specific need for developer workflows, justifying the count.

    Completeness3/5

    The tools comprehensively cover read operations for GitLab (commits, files, MRs, pipelines, code search) and Jira (issues, epics, sprints, attachments, comments). However, write operations (creating/updating MRs or issues) are absent, leaving a gap for lifecycle management.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      F
      maintenance
      A Model Context Protocol server that enables interaction with Jira's REST API using natural language commands, allowing users to manage Jira projects, issues, comments, and workflows through Claude Desktop and other MCP clients.
      10
      7
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      A read-only MCP server that provides AI agents structured access to Jira Cloud, enabling project listing, sprint overview, issue retrieval, and JQL search.
      1,667 npm
      MIT
    • A
      license
      B
      quality
      D
      maintenance
      MCP server providing access to JIRA data with relationship tracking, optimized payloads, and data cleaning for AI contexts. Supports both Jira Cloud and Server/Data Center.
      9
      18 npm
      MIT