glue
PlatformHow It WorksBenefitsPricingBlog
Log inGet Started
glue

The Product OS for engineering teams. Glue does the work. You make the calls.

hello@getglueapp.com

Product

  • Platform
  • How It Works
  • Benefits
  • Pricing
  • Stella

Solutions

  • Understand Your Code
  • Manage the Work
  • Grow the Product
  • For CTOs
  • For Engineering Leaders
  • For Product Managers

Resources

  • Blog
  • Guides
  • Glossary
  • Use Cases
  • Comparisons
  • About
  • Authors

© 2026 Glue. All rights reserved.

Privacy·Terms·Support

Stella

Agent orchestration for Glue. Agents observe your systems, create plans, execute workflows step by step, and produce artifacts — with human approval at every stage.

Agents are config, not code. One execution engine runs every workflow.

EventWebhook / Cron
ObserveAgent detects
PlanMatch playbook
ReviewUser approves
ExecuteStep by step
DeliverTo destination
Core Concept

The Pipeline

Every action in Glue follows one pipeline. An event triggers an agent, which creates an observation, which becomes a plan the user can approve. The system executes the plan step by step, producing artifacts at each stage.

Agents Observe

Observer agents watch your systems 24/7 — GSC rankings, Sentry errors, PostHog usage, Jira tickets, competitor websites, codebase changes. When something happens, they create an observation.

System Plans

The system matches each observation to a playbook — a predefined workflow that knows how to resolve it. It creates a plan: "here's what happened, here's what I recommend, here are the steps." The user never sees raw data.

Agents Execute

When the user approves, executor agents run each step — calling APIs, searching code, generating content, creating tickets. Each step produces an artifact the user can review, edit, or reject.

Key Insight

The user never sees an observation. They see a plan — the system already analyzed the situation and is presenting a recommended course of action with clear steps. The morning brief summarizes what agents found and did overnight.

Data Model

7 Entities

Stella has 7 first-class entities. Each maps to a database table prefixed with stella_. Together they represent the complete lifecycle from signal to output.

In Practice

Live Walkthrough

Follow a real event through the entire pipeline — from webhook to destination.

1
Sentry Webhook Fires

TypeError: Cannot read properties of undefined (reading 'map') — 47 events in 1 hour. Error Tracker agent receives the webhook.

2
Observation Created

stella_observations: source='sentry', title='TypeError spike — 47 events/hr', raw_data contains full stack trace, affected file, error count.

3
Plan Created

System matches to playbook 'Investigate Error' (#15). Creates stella_plans with severity='critical'. Step 0 (Detect) is pre-completed with the observation data as its artifact.

Artifact: analysis — error detection summary
Auto-approved
4
User Opens Product Health Page

Morning brief: 'TypeError spike in dashboard — 47 events. Agent has analyzed the error and traced it to DashboardPage.tsx:142.' Plan appears in the insight queue with [Investigate] button.

5
Step 1: Analyze Error

Agent reads the stack trace, searches codebase for the function, checks git blame for recent changes. Produces analysis: 'Null check missing on repos.map() — happens when project has no repositories.'

Artifact: analysis — root cause + affected users
Human review required
6
Step 2: Trace to Code

Agent walks the call graph: DashboardLayout → DashboardPage → RepoList.map(). Identifies the exact line and suggests fix: add optional chaining repos?.map().

Artifact: code — file:line reference + fix
Human review required
7
Step 3: Write Fix Spec

Agent generates a specification: add null guard at line 142, add empty state component, estimated 30 minutes.

Artifact: draft — fix specification (editable)
Human review required
8
Step 4: Create Ticket

Agent creates GLUE-451 in Jira. Priority: P1. Assigned: Sarah K. Sprint: 15 (current). Links to Sentry issue.

Artifact: confirmation — ticket created + links
Auto-approved
The Agents

Observer + Executor

Observer agents watch your systems and surface signals. Executor agents do the work — writing, analyzing, creating. Some agents do both.

Observer Agents

GSC Monitorobserverhaiku-4-5

Watches: Google Search Console rankings, CTR, indexing

Creates: Ranking drops, CTR anomalies, indexing issues

gsc_api
Error Trackerobserverhaiku-4-5

Watches: Sentry / PostHog error events

Creates: Error spikes, new exceptions, regressions

sentry_apiposthog_api
Usage Monitorobserverhaiku-4-5

Watches: PostHog analytics

Creates: Feature usage drops, funnel degradation

posthog_api
Competitor Watcherobserversonnet-4-6

Watches: Competitor websites (scheduled crawl)

Creates: New features, content updates, pricing changes

web_searchweb_scrape
Ticket Monitorobserverhaiku-4-5

Watches: Jira / Linear / ClickUp webhooks

Creates: New tickets, overdue tasks, blocked items

jira_apilinear_api
Security Scannerobserverhaiku-4-5

Watches: npm audit, Dependabot, GitHub alerts

Creates: CVEs, vulnerable packages

github_apinpm_audit

Executor Agents

Content Writerexecutorsonnet-4-6

Creates: Blog posts, articles, meta tags

web_searchcodebase_searchimage_gen
Code Analyzerexecutorsonnet-4-6

Creates: Error traces, root cause analysis, fix specs

codebase_searchsymbol_lookupcall_graph
Spec Writerexecutorsonnet-4-6

Creates: PRDs, technical specifications

codebase_searchticket_searchschema_lookup
Dev Plan Builderexecutorsonnet-4-6

Creates: Implementation plans, file-by-file breakdowns

codebase_searchcall_graphfile_analysis
Triage Agentbothhaiku-4-5

Watches: New tickets (via Ticket Monitor)

Creates: Classifications, priority assessments, sprint suggestions

jira_apicodebase_searchticket_search
SEO Optimizerexecutorhaiku-4-5

Creates: Meta tag rewrites, schema markup, SERP previews

gsc_apiweb_search
Capabilities

49 Tools

Tools are what agents use to interact with systems. Each playbook step declares which tools it needs — the execution engine provides only those. This is the permission boundary.

Read — Codebase (12 tools)

semantic_searchSearch code by meaning
search_symbolsFind functions/classes
get_method_bodyRead function source
get_type_definitionRead types/interfaces
get_config_constantsRead config values
find_api_endpointsList API routes
get_symbol_call_graphTrace caller→callee
search_featuresSearch discovered features
get_schemaDatabase table structure
get_file_structureList directory files
list_commitsGit commit history
get_file_historyFile ownership & authors

Read — External APIs (10 tools)

gsc_search_performanceGSC rankings
gsc_sitemapsGSC sitemaps
dataforseo_keywordsKeyword volumes
dataforseo_backlinksBacklink data
dataforseo_serpSERP results
posthog_trendsUsage trends
posthog_funnelFunnel analysis
posthog_retentionRetention cohorts
search_sentry_issuesError search
get_sentry_error_statsError metrics

Read — Synced Data (5 tools)

query_page_analyticsPage-level analytics
query_session_analyticsSession-level analytics
query_search_performanceSearch performance data
query_content_performanceContent performance data
query_funnel_analyticsFunnel analytics data

Write — Destinations (14 tools, 10 needed)

save_triage_detailsSave triage output
save_suggested_expertsSave expert suggestions
update_specificationUpdate spec document
save_agent_dataPersist agent output
create_jira_ticketComing
Create Jira ticket
update_jira_ticketComing
Update Jira ticket
create_github_prComing
Create GitHub PR
publish_to_cmsComing
Publish to CMS
submit_indexnowComing
Submit IndexNow ping
submit_gsc_reindexComing
Request GSC reindex
publish_to_hashnodeComing
Publish to Hashnode
send_slack_messageComing
Send Slack message
create_redirectComing
Create URL redirect
generate_imageComing
Generate image asset

How steps get their tools

Research step
web_searchgsc_search_performancesemantic_search
Write step
semantic_searchget_method_bodyweb_search
Publish step
publish_to_cmssubmit_indexnow

The execution engine creates an MCP server with ONLY the declared tools. The LLM cannot call tools outside this set.

39built
10needed
1engine
Playbook Library

33 Built-in Playbooks

Each playbook defines a workflow — ordered steps, tools per step, artifact types, and approval chain. Users can also create custom playbooks.

SEO & Content (12)

Product Health (8)

Competitive (3)

Roadmaps (6)

Releases (3)

Architecture Principle

Agents Are Config, Not Code

Every agent is a row in the database. The execution engine reads the config and runs it. To add a new agent, insert rows — no TypeScript required.

Before — Hardcoded

// ticket-triage-agent-sdk.ts
// 400+ lines of custom code PER agent

function buildTriagePrompt(ticket) {
  return `# TICKET ANALYSIS
  ## Phase 1: Triage (MANDATORY)
  Use ticket-triage skill...
  ## Phase 2: Specification
  Use spec-generation skill...
  `;
}

// Every agent = a new TypeScript file
// Every prompt change = code deploy
// No user customization possible

After — Config in DB

-- stella_playbooks
INSERT INTO stella_playbooks
  (name, page_key, model)
VALUES
  ('Triage New Ticket', 'roadmaps', 'haiku-4-5');

-- stella_playbook_steps
INSERT INTO stella_playbook_steps
  (playbook_id, step_index, name,
   agent_prompt, tools, artifact_type)
VALUES
  (26, 0, 'Classify',
   'Classify ticket {observation.title}...',
   '{codebase_search,ticket_search}',
   'classification');

-- One engine runs ALL playbooks
-- Prompt changes = DB update (no deploy)
-- Users create custom playbooks in UI
System Architecture

Java + TypeScript

The execution engine runs in Java. The UI runs in Next.js. One request flow connects them.

Next.js (boostr-web)

TypeScript / React / App Router

PagesSEO, Health, Competitive, Roadmaps, Releases, Agents
Workflow modal14 artifact renderers
Morning briefDaily agent summary per page
Auth + sessionNextAuth, RBAC, project scoping
API routesThin proxy to Java
SSE displayPipe Java stream to browser

Java (Stella)

Spring Boot / PostgreSQL / MCP

AgentExecutorAgentic loop (observe → plan → execute)
SessionOrchestratorMulti-agent routing
ToolRegistry49 tools
Integration clientsJira, Linear, ClickUp, Sentry, GitHub
30+ DAOsstella_ tables
Webhook receiversCron scheduler
MCP serverTool serving over stdio/HTTP
SSE streamingReal-time artifact delivery

Request Flow

1
User clicks actionFix Ranking Drop on SEO page
2
Next.js proxiesPOST /api/stella/plans/start (auth check, then forward to Java)
3
Java orchestratesSessionOrchestrator loads playbook, AgentExecutor runs each step with tools
4
Artifacts stream backSSE events flow: Java → Next.js → Browser → Workflow modal renders
5
User reviewsApprove, edit, or reject. Auto-approved steps skip this.

Why Java, Not Claude Agent SDK

CapabilityJava StellaClaude Agent SDK
Agentic loop
AgentExecutor
Built-in
Tool persistence
Logs every call
Not built-in
Cost tracking
Per-iteration tracking
Not built-in
Multi-agent routing
SessionOrchestrator
Single-agent
Integration clients
5 built (Jira, Linear, ClickUp, Sentry, GitHub)
Must build

The Agent SDK is great for single-agent tasks. Java Stella is the orchestration layer that coordinates multiple agents, persists every tool call, tracks costs, and routes requests across sessions.

Migration Path

How Existing Agents Fit

The current agent codebase maps directly to Stella entities. The intelligence stays — the orchestration changes.

Current CodeStella EquivalentStatus
ticket-triage-agent-sdk.tsstella_playbooks #26 (Triage New Ticket)refactor
seo-analysis-agent-sdk.tsstella_playbooks #3-14 (12 SEO playbooks)refactor
feature-enrichment-agent-sdk.tsstella_playbooks #1 (Enrich Feature)refactor
gap-analysis-agent-sdk.tsstella_playbooks #23-25 (Competitive)refactor
comprehensive-gap-agent.tsMerged into playbook #23refactor
getAgentToolConfig()stella_agents.tools + stella_playbook_steps.toolsrefactor
task_types tablestella_playbooks + stella_playbook_stepsrefactor
insights tablestella_observations → stella_plansrefactor
action_tasks tablestella_plansrefactor
task_steps tablestella_stepsrefactor
task_artifacts tablestella_artifactsdone
agents tablestella_agentsdone
agent_runs tableLinked via stella_observations.agent_run_iddone
agent_templates tablestella_playbooks (is_system=true)refactor
skills tableNot needed — tools defined on steps directlydone
Workflow modal (workflow-modal.tsx)No change — reads from stella_artifactsdone
Mock pages (5 pages)Real pages reading from stella_plans APIrefactor
Extensibility

Custom Playbooks

Users build their own playbooks on the Agents page. Same tables, same engine, same artifact renderers. Define a trigger, choose tools, set approval chains.

1Define trigger— "When a PR with 500+ lines is opened"
2Add steps— "Analyze PR → Suggest split → Notify author"
3Set approval— "Auto-approve analysis, human-review suggestions"
4Activate— Same engine, same artifact renderers, zero code
FAQ

Questions Your Team Will Ask

Answers for engineers, PMs, and anyone working with Stella.

7 tables. One engine. Unlimited workflows.

stella_agents observe. stella_observations record signals. stella_playbooks define resolutions. stella_plans present them to users. stella_steps execute. stella_artifacts produce output. Users approve. Destinations receive.

7Tables
33Built-in Playbooks
14Artifact Renderers
1Execution Engine