Repository navigation
Expand file tree
/
Copy pathbuggregator.yaml.example
More file actions
235 lines (214 loc) · 13 KB
/
Copy pathbuggregator.yaml.example
File metadata and controls
235 lines (214 loc) · 13 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
# Buggregator configuration file.
#
# Place as buggregator.yaml next to the binary, or specify with --config flag.
# All values support environment variable substitution: ${VAR} and ${VAR:default}
#
# Every setting has a sensible default — this file is entirely optional.
# HTTP server
server:
addr: ${HTTP_ADDR::8000}
cors_origins: # Allowed CORS origins. Default: ["*"] (all origins).
- "*" # Use specific origins to restrict: ["http://localhost:3000"]
# Database
database:
driver: sqlite # Only SQLite supported for now
dsn: ${DATABASE_DSN::memory:} # ":memory:" (default), "data.db", or full path
max_open_conns: ${DATABASE_MAX_OPEN_CONNS:4} # SQLite connection pool size. 1 serializes the UI with ingestion.
# Needs journal_mode(WAL) in the DSN to be worth anything.
# Attachment storage (SMTP attachments, HTTP dump files)
storage:
mode: ${STORAGE_MODE:memory} # "memory" (default, lost on restart) or "filesystem"
path: ${STORAGE_PATH:./storage} # Directory for filesystem mode
# Retention: delete old events in the background (every minute). Unset keeps
# everything. Pinned events are never deleted and do not count towards
# max_events. Attachments are deleted with their events.
retention:
max_age: ${RETENTION_MAX_AGE:} # e.g. "24h", "7d"; empty or 0 = keep
max_events: ${RETENTION_MAX_EVENTS:0} # keep at most this many events in total; 0 = no cap
# Sentry: JS stack frames carry only a URL; the server fetches that file to show
# the code. Ingestion is open even with auth, so by default only loopback and
# private hosts (dev server, docker service, ddev) are fetched.
sentry:
source_fetch: ${SENTRY_SOURCE_FETCH:private} # private | any | off
# Release a caught email to a real inbox (button on an SMTP message). Off while
# host is empty. Nothing is forwarded on its own: a message goes out only on a
# release request. Without auth, release works only from this machine
# (localhost) and only to allowed_recipients, which must then be set, each
# with an exact domain ("*@mycompany.com", not "*" or "*@*"): a proxy on this
# host makes outside requests look local. In Docker, from another
# machine or behind a proxy turn auth on (then the list is optional).
# Keep the password in SMTP_RELAY_PASSWORD or a file.
smtp_relay:
host: ${SMTP_RELAY_HOST:} # e.g. smtp.gmail.com, or mailpit for a local test
port: ${SMTP_RELAY_PORT:0} # 0 = 587 for starttls, 465 for tls, 25 for none
tls: ${SMTP_RELAY_TLS:starttls} # starttls | tls | none
username: ${SMTP_RELAY_USERNAME:} # empty = no login
password_file: ${SMTP_RELAY_PASSWORD_FILE:} # or set SMTP_RELAY_PASSWORD
from: ${SMTP_RELAY_FROM:} # envelope sender; empty = username when it is an address (Gmail, SES want the account), else the message's From
allowed_recipients: [] # e.g. ["*@mycompany.com", "me@gmail.com"]; env SMTP_RELAY_ALLOWED_RECIPIENTS (comma list); required without auth, empty with auth = any
# Broadcast events (module broadcast): Buggregator answers the Pusher HTTP API
# (POST /apps/{app_id}/events, /batch_events) that Laravel Reverb, Pusher and
# soketi clients call. Point REVERB_HOST/REVERB_PORT (or PUSHER_HOST/PUSHER_PORT)
# at Buggregator; each event becomes a card. Without upstream the app gets {}
# and browsers get nothing; with upstream the same request also goes to the
# real server and the app gets its answer. Change only the server-side
# REVERB_HOST/REVERB_PORT: keep VITE_REVERB_* (the browsers' WebSocket) on
# Reverb, as Laravel's .env derives them from REVERB_HOST by default. Only
# POSTs are caught: channel queries (GET /apps/{id}/channels) are not.
broadcast:
upstream: ${BROADCAST_UPSTREAM:} # e.g. http://127.0.0.1:8080 (Reverb). Empty = capture only
# apps: # check request signatures; env BROADCAST_APPS="key:secret,..."
# - key: app-key
# secret: app-secret
# UI event lists: how much /api/events and /api/events/preview return
ui:
default_limit: ${UI_DEFAULT_LIMIT:} # Events returned when the request has no limit param. Empty or 0 = unlimited
max_limit: ${UI_MAX_LIMIT:5000} # Ceiling for an explicit ?limit=. 0 = no ceiling
default_window: ${UI_DEFAULT_WINDOW:} # Default time window, e.g. "1h", "24h", "7d". Empty = no window
# New-version check: asks GitHub (api.github.com, releases/latest) for the newest
# server release; the UI shows a notice when it is newer. Off = no network call.
update_check:
enabled: ${UPDATE_CHECK_ENABLED:true}
interval: ${UPDATE_CHECK_INTERVAL:12h} # Go duration, at least 1h
# Update from the UI (binary installs). auto: allowed unless the server
# listens beyond 127.0.0.1 without auth; on: allowed then too; off: never.
# Docker, Kubernetes and a read-only install dir get the command to run instead.
# The download is checked against checksums.txt and its ed25519 signature.
update:
self_update: ${UPDATE_SELF:auto}
# Xdebug function traces (xdebug.trace_format=1), uploaded over HTTP:
# curl -H 'X-Buggregator-Event: xdebug-trace' --data-binary @trace.xt.gz http://127.0.0.1:8000/
# Optional headers: X-Buggregator-Project, X-Buggregator-Trace-Url (the request
# traced), X-Buggregator-Trace-Name. Files are kept as sent (gzip or not) under
# storage.path/xdebug-traces in filesystem mode, in a temp dir otherwise.
xdebug_trace:
max_upload: ${XDEBUG_TRACE_MAX_UPLOAD:100MB} # request body (KB/MB/GB)
max_size: ${XDEBUG_TRACE_MAX_SIZE:256MB} # uncompressed; a longer trace is cut there
keep: ${XDEBUG_TRACE_KEEP:20} # traces kept; the oldest unpinned go first
# Prometheus metrics
metrics:
enabled: ${METRICS_ENABLED:false} # Set to true to expose Prometheus metrics
addr: ${METRICS_ADDR:} # Separate metrics server address (e.g., ":9090")
# Leave empty to serve /metrics on the main HTTP server
# Authentication via OAuth2/OIDC providers.
# Supported providers: auth0, google, github, keycloak, gitlab, oidc (generic OIDC).
# For OIDC-compatible providers (auth0, google, keycloak, gitlab, oidc), endpoints are
# auto-discovered from provider_url via .well-known/openid-configuration.
# For GitHub, endpoints are hardcoded (no OIDC support).
# auth:
# enabled: false
# provider: ${AUTH_PROVIDER:oidc} # Provider type
# provider_url: ${AUTH_PROVIDER_URL:} # OIDC issuer URL (e.g., https://xxx.us.auth0.com)
# client_id: ${AUTH_CLIENT_ID:} # OAuth2 client ID
# client_secret: ${AUTH_CLIENT_SECRET:} # OAuth2 client secret
# callback_url: ${AUTH_CALLBACK_URL:} # Callback URL (e.g., http://localhost:8000/auth/sso/callback)
# scopes: ${AUTH_SCOPES:openid,email,profile} # Comma-separated scopes
# jwt_secret: ${AUTH_JWT_SECRET:} # Secret for signing internal JWT tokens (required)
# TCP servers
tcp:
smtp:
addr: ${SMTP_ADDR::1025}
monolog:
addr: ${MONOLOG_ADDR::9913}
var-dumper:
addr: ${VAR_DUMPER_ADDR::9912}
# MCP server: lets AI agents (Claude Code, Codex, Cursor...) read events, Sentry
# errors, dumps and profiles, and delete events. Off by default.
# transport: socket — a Unix socket on this machine; the agent starts `buggregator mcp`
# as its MCP command, which connects to the socket. No auth: keep the socket local.
# transport: http — remote access on addr; set auth_token, or anyone who can reach
# the port can use the tools.
# mcp:
# enabled: ${MCP_ENABLED:false}
# transport: ${MCP_TRANSPORT:socket} # socket or http
# socket_path: ${MCP_SOCKET_PATH:/tmp/buggregator-mcp.sock} # For transport=socket
# addr: ${MCP_ADDR::8001} # For transport=http (all interfaces)
# auth_token: ${MCP_AUTH_TOKEN:} # Bearer token for transport=http. Empty = no auth
# mock_write: ${MCP_MOCK_WRITE:true} # mock_rule_create/update/delete/from_request, mock_import_openapi, mock_group_delete; false = read and preview only
# Xdebug proxy: PHP/Xdebug -> Buggregator -> IDE, plus MCP tools (xdebug_sessions,
# xdebug_location, xdebug_stack, xdebug_variables, xdebug_property, xdebug_eval) on
# the same paused session; xdebug_eval runs PHP code, only with allow_eval.
# With allow_control an agent also drives a session that has no IDE
# (xdebug_step, xdebug_continue, xdebug_stop), waits for a pause (xdebug_wait) and
# sets breakpoints (xdebug_breakpoint_set, xdebug_breakpoint_remove; listing them,
# xdebug_breakpoints, is always on). Breakpoints are saved in the database and sent
# to every session Buggregator drives; a condition is PHP code and needs allow_eval.
# Off by default. MCP tools need mcp.enabled.
# mode: proxy — relay to the IDE; no IDE = the script runs on without a debugger.
# auto — relay when the IDE answers, else Buggregator drives the session
# (step, continue, stop in the UI). standalone — never dial the IDE.
# While an IDE is attached it drives; Buggregator only reads.
# Same host: point Xdebug at addr (xdebug.client_port=9003) and move the IDE's
# debug port to ide_addr (PhpStorm: Settings > PHP > Debug > Debug port 9004).
# xdebug:
# enabled: ${XDEBUG_PROXY_ENABLED:false}
# addr: ${XDEBUG_PROXY_ADDR:127.0.0.1:9003} # Where Xdebug connects
# ide_addr: ${XDEBUG_PROXY_IDE_ADDR:127.0.0.1:9004} # Where the IDE listens
# allow_eval: ${XDEBUG_PROXY_ALLOW_EVAL:false} # Let MCP clients evaluate PHP expressions
# allow_control: ${XDEBUG_PROXY_ALLOW_CONTROL:true} # MCP step/continue/stop/wait; false = read only
# mode: ${XDEBUG_PROXY_MODE:auto} # proxy, auto or standalone
# pause_on_first_line: ${XDEBUG_PROXY_PAUSE_ON_FIRST_LINE:false} # With no IDE: stop at the first line
# idle_timeout: ${XDEBUG_PROXY_IDLE_TIMEOUT:10m} # With no IDE: release a session paused this long (a Go duration: 90s, 10m); 0 = never
# Enable/disable modules. All enabled by default.
# Set to false to disable a module entirely (TCP port won't bind, events ignored).
# Can also be controlled via CLIENT_SUPPORTED_EVENTS env (comma-separated list).
modules:
sentry: true
ray: true
var-dump: true
inspector: true
monolog: true
smtp: true
sms: true
broadcast: true
http-dump: true
profiler: true
xdebug-trace: true
# Webhooks — send HTTP POST notifications when events are received.
# Each webhook fires for a specific event type, or "*" for all events.
# Payload: { uuid, type, payload, timestamp, project }
# webhooks:
# - event: "*" # Fire on any event type
# url: https://slack.example.com/webhook
# headers:
# Authorization: "Bearer token123"
# verify_ssl: false # Default: false
# retry: true # Retry up to 3 times with exponential backoff (default: true)
# - event: sentry # Fire only on Sentry events
# url: https://pagerduty.example.com/alert
# - event: smtp # Fire only on SMTP events
# url: https://my-app.example.com/email-hook
# Mock server — requests to http://<host>:8000/mock<path> get the response of
# the first matching rule and are captured as HTTP dumps. Rules defined here are
# read-only in the UI and come before the rules made in the UI.
# Path forms: /users/me (exact), /users/{id} (one segment), /files/* (the rest).
# Unmatched requests get 404 with a JSON error and are still captured.
# mocks:
# - method: GET # GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS or ANY (default)
# path: /users/{id}
# status: 200 # default 200
# headers: { Content-Type: application/json }
# body: '{"id": 1, "name": "Jane"}'
# delay: 150ms # optional, at most 30s
# project: default # optional; empty = every project
# - method: GET # a dynamic response: request values and fake data
# path: /orders/{id}
# template: true # body and header values are Go templates
# seed: 42 # optional; not 0 = the same fake data on every request
# headers: { Content-Type: application/json }
# body: '{"id": {{ path "id" }}, "customer": {{ fake "name" | json }}, "total": {{ fake "price" 5 500 }}}'
# OpenAPI 3.0/3.1 specs (YAML or JSON) turned into read-only mock rules at start:
# one rule per operation; the spec's example as the body, else a template with
# fake data built from the schema. Paths get servers[0].url's path in front.
# mocks_openapi:
# - file: ./openapi.yaml
# project: shop # optional; empty = every project
# group: shop-api # optional; default openapi:<project>:<info.title>
# base_path: "" # optional; "" = no prefix
# Pre-defined projects. The "default" project always exists.
# Additional projects can be defined here.
# projects:
# - key: my-app
# name: My Application
# - key: staging
# name: Staging Environment