githubFixer/.env.example at main · mando222/githubFixer · GitHub

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
# ── Required ──────────────────────────────────────────────────────────────────

# Linear — get API key from Linear > Settings > API > Personal API Keys
LINEAR_API_KEY=lin_api_...
LINEAR_TEAM_ID=

# Anthropic API key — required when AGENT_BACKEND=anthropic (the default)
ANTHROPIC_API_KEY=sk-ant-...

# ── Agent backend ──────────────────────────────────────────────────────────────
# "anthropic" — direct Anthropic API (default, no CLI required)
# "codex"     — OpenAI Codex CLI (requires: npm install -g @openai/codex + OPENAI_API_KEY)
AGENT_BACKEND=anthropic

# Per-agent overrides — leave blank to use AGENT_BACKEND above
# Example: run the coder via Codex while everything else uses Anthropic
# CODER_AGENT_BACKEND=codex
# TESTER_AGENT_BACKEND=
# REVIEWER_AGENT_BACKEND=
# ANALYZER_AGENT_BACKEND=
# PLANNER_AGENT_BACKEND=
# SPEC_WRITER_AGENT_BACKEND=
# SPEC_REVIEWER_AGENT_BACKEND=
# GITHUB_AGENT_BACKEND=

# ── Anthropic models (optional — defaults shown) ───────────────────────────────
CODING_AGENT_MODEL=claude-sonnet-4-6
TESTER_AGENT_MODEL=claude-sonnet-4-6
REVIEWER_AGENT_MODEL=claude-sonnet-4-6
SPEC_WRITER_AGENT_MODEL=claude-sonnet-4-6
GITHUB_AGENT_MODEL=claude-haiku-4-5-20251001
ANALYZER_AGENT_MODEL=claude-haiku-4-5-20251001
PLANNER_AGENT_MODEL=claude-haiku-4-5-20251001
SPEC_REVIEWER_AGENT_MODEL=claude-haiku-4-5-20251001

# Max output tokens per Anthropic API call (default: 16384)
# MAX_TOKENS_PER_AGENT=16384

# ── Codex CLI backend (optional) ──────────────────────────────────────────────
# Only needed if AGENT_BACKEND=codex or any *_AGENT_BACKEND=codex
# OPENAI_API_KEY=sk-...   # not needed if you use `codex login`
# CODEX_CODER_MODEL=o4-mini
# CODEX_TESTER_MODEL=o4-mini
# CODEX_REVIEWER_MODEL=o4-mini
# CODEX_ANALYZER_MODEL=o4-mini
# CODEX_PLANNER_MODEL=o4-mini
# CODEX_SPEC_WRITER_MODEL=o4-mini
# CODEX_SPEC_REVIEWER_MODEL=o4-mini
# CODEX_GITHUB_MODEL=o4-mini
# CODEX_TIMEOUT_SECONDS=600

# ── Ollama (optional — zero-cost local inference for tool-free agents) ─────────
# Ollama runs locally and is used for Planner and/or Spec Reviewer.
# Falls back to Anthropic API automatically if Ollama is unreachable.
# USE_OLLAMA_FOR_PLANNER=false
# USE_OLLAMA_FOR_SPEC_REVIEWER=false
# OLLAMA_MODEL=qwen2.5:14b
# OLLAMA_BASE_URL=http://localhost:11434

# ── Concurrency ────────────────────────────────────────────────────────────────
MAX_CONCURRENT_ISSUES=3
# MAX_CONCURRENT_TESTERS=5
# MAX_CONCURRENT_PLANNERS=5

# ── Timeouts & retry limits ────────────────────────────────────────────────────
# ISSUE_TIMEOUT_SECONDS=1800
# PLANNING_TIMEOUT_SECONDS=900
# MAX_REMEDIATION_CYCLES=3
# MAX_REVIEW_CYCLES=2

# ── Linear state IDs (optional — enables richer status tracking) ───────────────
# Find these in Linear > Settings > Workflow
# LINEAR_IN_PROGRESS_STATE_ID=
# LINEAR_IN_REVIEW_STATE_ID=
# LINEAR_DONE_STATE_ID=
# LINEAR_NEEDS_CLARIFICATION_STATE_ID=

# ── mempalace (optional — persistent cross-run memory) ─────────────────────────
# Enables: codebase analysis caching, prior decision injection, failure pattern memory.
# Setup:
#   pip install git+https://github.com/milla-jovovich/mempalace
#   mempalace init ~/.mempalace
# MEMPALACE_ENABLED=false
# MEMPALACE_PALACE_PATH=~/.mempalace/palace