career-coach
claudeTracks Soham's own growth across every project using persistent cross-project memory - not project status. Install once at --scope user.
1 agent
DevOps/developer/tester subagent team, guardrail hooks, project bootstrap skill, and GitLab MCP wiring. Install per project at --scope project.
Plugin signoff from marketplace soham-team-kit in Data-eng15/Signoff. Commands read from the repo's manifest — 1 month ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.
bootstrap-projectScaffolds the standard project structure (CLAUDE.md, docs/, .gitlab-ci.yml, Dockerfile, issue/MR templates) in the current empty repo, using this plugin's configured parameters. Use once, when a new project repo is created.
checkpoint-reviewInterviews Soham on a subagent's deliverable (SRS, architecture, design, or sprint plan) before it's marked reviewed - reads the deliverable's Review Questions, asks them one at a time, evaluates his answers against the actual reasoning in the document, and only logs the stage as reviewed once he's
project-intakeRuns before pm-coordinator, in the main session where AskUserQuestion is available - asks Soham the process-level questions pm-coordinator can't ask itself once it's running as a delegated subagent (sprint length, capacity, GitLab issue creation preference, first pass vs revision), then hands pm-coo
standupGenerates an end-of-day standup update (yesterday/today/blockers) from actual GitLab activity - commits, MR status, and issue moves. Use at end of day before the real standup.
ui-requirements-gateRuns before ui-ux-designer, in the main session where AskUserQuestion is available - asks Soham the constraints ui-ux-designer can't ask about itself once it's running as a delegated subagent (fixed brand constraints not in the SRS, screens explicitly out of scope this pass, whether a connected desi
architecture-criticAdversarial review of docs/architecture.md before the architecture checkpoint interview - hunts specifically for mistake classes that are expensive to catch late (solved-the-wrong-problem, data-model decisions costly to reverse, coupling that blocks scaling, circular dependencies). A different job t
automation-testerWrites and maintains the automated test suite (unit, integration, and end-to-end) so regressions get caught by CI instead of a human re-checking by hand every time. Use after a story is verified by manual-tester, or when test coverage is falling behind.
backend-developerImplements API and backend features and fixes bugs, based on a linked GitLab issue's acceptance criteria. Use for any work under /src/backend or equivalent API code.
business-analystTurns a raw client or stakeholder ask into a structured SRS (Software Requirements Specification). Use at the start of any new project or major feature, before any sprint planning or architecture work begins.
cloud-engineerAnalyzes technical feasibility, cloud provider choice, and rough infra cost for a raw project ask, before any SRS exists. This is a feasibility read, not a build - separate from devops-engineer, which provisions real infra once architecture is decided. Only invoke when Soham explicitly asks for clou
devops-engineerTwo modes. SETUP runs first in a brand-new repo, before architecture exists - creates the empty environment/CI shell so research and the rest of the pipeline have somewhere to run. PROVISION runs later, once docs/architecture.md exists, and stands up whichever cloud provider that document specifies
finance-analystAssesses whether a raw project idea is worth building - business problem solved, rough build cost vs. value, go/no-go recommendation - before any SRS is written. Only invoke when Soham explicitly asks for finance-analyst by name.
frontend-developerImplements UI features and fixes frontend bugs, based on a linked GitLab issue's acceptance criteria. Use for any work under /src/frontend or equivalent UI code.
manual-testerManually verifies a completed feature or fix against its issue's Definition of Done by actually exercising the app - not just reading the diff. Only invoke when Soham explicitly asks for manual-tester by name.
mcp-provisionerReads docs/research-discover.md after DISCOVER mode completes and recommends which project-specific MCPs each downstream agent (business-analyst, system-architect, devops-engineer, frontend-developer, backend-developer, etc.) should have wired up for this project. Does not touch global/always-on MCP
pm-coordinatorBreaks a verified SRS into epics, stories, and GitLab issues with acceptance criteria, and organizes them into sprint milestones. Use after the SRS is approved and before development starts.
researchRuns project research in three modes - DISCOVER (scope a raw idea before cloud-engineer runs feasibility), INVESTIGATE (answer a specific technical/market/prior-art question mid-sprint), CHALLENGE (actively try to kill a requirement - is it necessary or over-engineering). State which mode you want w
security-auditorAudits Signoff's own harness surface — hooks, agent tool grants, MCP config, scripts — for security issues, as distinct from auditing the client project's generated code. Not part of the research→business-case→SRS→architecture→design→sprint-plan pipeline and not gated by require-review-gate.sh. Run
system-architectDesigns the technical architecture from a verified SRS and story breakdown, before devops or development starts. Use once the SRS is approved and stories exist.
ui-ux-designerDesigns the actual user experience - flows, screens, component inventory, design system - from the SRS and the approved architecture, before pm-coordinator sizes stories or frontend-developer builds anything. Use once docs/architecture.md exists and has been through checkpoint-review.
gitlabruns: REPLACE_WITH_YOUR_GITLAB_MCP_SERVER_COMMAND
"${CLAUDE_PLUGIN_ROOT}"/scripts/block-main-push.sh
"${CLAUDE_PLUGIN_ROOT}"/scripts/block-ai-attribution.sh
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/emit_event.py
"${CLAUDE_PLUGIN_ROOT}"/scripts/lint-on-save.sh
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/capture-agent-tool-response.sh
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/emit_event.py
"${CLAUDE_PLUGIN_ROOT}"/scripts/require-review-gate.sh
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/emit_event.py
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/emit_event.py
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/open-completed-doc.sh
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/emit_event.py
"${CLAUDE_PLUGIN_ROOT}"/scripts/observability/start-dashboard.sh
Tracks Soham's own growth across every project using persistent cross-project memory - not project status. Install once at --scope user.
1 agent
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.
1 MCP server
Build powerful, stateful CLI interfaces for any GUI application using the cli-anything harness methodology.
5 commands
Automates browser interactions for web testing, form filling, screenshots, and data extraction
12 skills
Use Alibaba Open Code Review from local Codex.
2 skills
Claude Code skill for creating and presenting Slidev slidedecks
1 skill
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.