forked from HAL9000/cleveragents-core
freemo
e5f75c5c83
- Remove maximum cap (16) on CA_MAX_PARALLEL_WORKERS in resources.yaml - Can now be set to any positive value (32, 64, etc.) - Only minimum validation remains (must be > 0) - Remove dynamic backpressure/throttling from implementation-orchestrator - Dispatch always runs at full configured speed - Resource monitoring remains for visibility only - No automatic reduction of slots_available based on failures - Convert system-watchdog from auto-degradation to monitoring + suggestions - Renamed DEGRADATION_THRESHOLDS to HEALTH_THRESHOLDS - Removed apply_system_degradation() and check_degradation_recovery() - Changed findings to include suggestions instead of actions - Watchdog now reports issues with fix recommendations - No automatic throttling or pausing of agents The system now operates at maximum configured speed at all times, with the watchdog providing diagnostic insights when issues arise.
139 lines
3.9 KiB
YAML
139 lines
3.9 KiB
YAML
# CleverAgents Resource Management Configuration
|
|
# Controls parallelism, timeouts, and resource allocation
|
|
|
|
resources:
|
|
# Maximum parallel workers
|
|
max_parallel_workers:
|
|
default: 4
|
|
min: 1
|
|
# No maximum - can be set to any positive value (e.g., 32, 64, etc.)
|
|
env_var: CA_MAX_PARALLEL_WORKERS
|
|
|
|
# Worker allocation tiers
|
|
# Based on fraction of max_parallel_workers
|
|
worker_allocation:
|
|
# Full tier (N) - Maximum throughput
|
|
implementation-orchestrator: 1.0
|
|
|
|
# Half tier (N/2) - Moderate parallelism
|
|
continuous-pr-reviewer: 0.5
|
|
|
|
# Quarter tier (N/4) - Limited parallelism to prevent scope explosion
|
|
uat-tester: 0.25
|
|
bug-hunter: 0.25
|
|
test-infra-improver: 0.25
|
|
|
|
# Singleton agents (no parallelism)
|
|
architect: 0
|
|
epic-planner: 0
|
|
human-liaison: 0
|
|
agent-evolver: 0
|
|
architecture-guard: 0
|
|
spec-updater: 0
|
|
backlog-groomer: 0
|
|
docs-writer: 0
|
|
timeline-updater: 0
|
|
project-owner: 0
|
|
system-watchdog: 0
|
|
|
|
# Timeout configurations (in seconds)
|
|
timeouts:
|
|
# Agent operation timeouts
|
|
agent_heartbeat: 600 # 10 minutes
|
|
work_claim_hours: 2 # Work claim duration
|
|
bash_command_default: 120 # 2 minutes
|
|
bash_sleep_multiplier: 1.5 # Sleep timeout = sleep_duration * multiplier
|
|
|
|
# Human response timeouts (hours)
|
|
human_response:
|
|
needs_feedback_pr: 48 # 2 days
|
|
question_on_issue: 24 # 1 day
|
|
blocked_on_human: 72 # 3 days
|
|
architecture_approval: 48 # 2 days
|
|
verification_request: 24 # 1 day
|
|
|
|
# API timeouts (milliseconds)
|
|
forgejo_api: 30000 # 30 seconds
|
|
git_operations: 300000 # 5 minutes
|
|
test_execution: 900000 # 15 minutes
|
|
|
|
# Queue management
|
|
queue_limits:
|
|
max_depth_multiplier: 3 # max_queue = workers * multiplier
|
|
priority_queue_size: 10 # High priority items to keep ready
|
|
|
|
# Retry policies
|
|
retry_policies:
|
|
# Network errors (API, git)
|
|
network_errors:
|
|
max_attempts: 3
|
|
backoff: exponential # 1s, 2s, 4s
|
|
base_delay_seconds: 1
|
|
max_delay_seconds: 60
|
|
|
|
# Git conflicts
|
|
git_conflicts:
|
|
max_attempts: 2
|
|
strategy: rebase_and_retry
|
|
cleanup_on_failure: true
|
|
|
|
# Test failures
|
|
test_failures:
|
|
max_attempts: 3
|
|
strategy: escalate_tier # codex -> sonnet -> opus
|
|
|
|
# API rate limits
|
|
rate_limits:
|
|
max_attempts: 5
|
|
backoff: fixed
|
|
delay_seconds: 60 # Wait 1 minute between attempts
|
|
|
|
# Resource pressure thresholds
|
|
pressure_thresholds:
|
|
# Worker utilization
|
|
high_utilization_percent: 90
|
|
critical_utilization_percent: 95
|
|
|
|
# Failure rates (triggers backpressure)
|
|
acceptable_failure_rate: 0.2 # 20%
|
|
high_failure_rate: 0.5 # 50%
|
|
critical_failure_rate: 0.8 # 80%
|
|
|
|
# Queue depth
|
|
queue_warning_threshold: 0.8 # 80% of max
|
|
queue_critical_threshold: 1.0 # 100% of max
|
|
|
|
# Monitoring intervals (seconds)
|
|
monitoring:
|
|
health_signal_interval: 600 # 10 minutes
|
|
worker_status_check: 60 # 1 minute
|
|
session_validation: 300 # 5 minutes
|
|
resource_cleanup: 300 # 5 minutes
|
|
|
|
# Cleanup policies
|
|
cleanup:
|
|
temp_dir_prefix: /tmp/cleveragents-
|
|
max_temp_dir_age_hours: 24
|
|
clean_on_exit: true
|
|
preserve_on_failure: false # Set true for debugging
|
|
|
|
# Session management
|
|
sessions:
|
|
naming_pattern: "[{tag}] {type}: {target}"
|
|
tag_prefixes:
|
|
implementation: AUTO-IMP
|
|
pr_review: AUTO-REV
|
|
uat_test: AUTO-UAT
|
|
bug_hunt: AUTO-BUG
|
|
test_infra: AUTO-INF
|
|
architecture: AUTO-ARCH
|
|
planning: AUTO-EPIC
|
|
human: AUTO-HUMAN
|
|
evolution: AUTO-EVLV
|
|
guard: AUTO-GUARD
|
|
spec: AUTO-SPEC
|
|
backlog: AUTO-BLOG
|
|
docs: AUTO-DOCS
|
|
timeline: AUTO-TIME
|
|
owner: AUTO-OWNR
|
|
watchdog: AUTO-WDOG |