Skip to content

fix(agent): raise per-response token budget so tool calls aren't truncated #138

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
bararchy merged 1 commit into from
Jun 29, 2026
Merged

fix(agent): raise per-response token budget so tool calls aren't truncated #138

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion examples/agent.cr
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -536,7 +536,11 @@ end
# Entry point
# ----------------------------------------------------------------------------
model_dir = ARGV[0]?
max_tokens = (ARGV[1]? || "512").to_i
# Per-response generation ceiling. This is a safety bound, not a target — the
# model stops itself on <|im_end|>. It must be large enough that a tool call
# writing a whole file isn't truncated before its closing </tool_call> tag
# (which would make the call unparseable). KV cache grows ~linearly with it.
max_tokens = (ARGV[1]? || ENV["SHAINET_AGENT_MAX_TOKENS"]? || "4096").to_i
unless model_dir && Dir.exists?(model_dir)
STDERR.puts "Usage: agent <model-dir> [max_tokens]"
STDERR.puts " (point at an already-downloaded model, e.g. ~/models/Qwen3-Coder-30B-A3B-Instruct)"
Expand Down
Loading