feat(runtime): route large tool outputs through workshop to protect parent context (closes #548)

Tool outputs (read_file, grep_files, exec_shell, fetch_url, web_search) that
exceed a configurable token threshold are now intercepted before they reach
the parent context. A structured synthesis header replaces the raw blob; the
full content is stored in the workshop variable `last_tool_result` for later
`promote_to_context` retrieval.

Key changes:
- New `crates/tui/src/tools/large_output_router.rs`: `LargeOutputRouter`,
  `WorkshopConfig`, `WorkshopVariables`, `RouteDecision`, token estimator,
  synthesis-prompt builder, and wrap_synthesis helper. Full unit-test suite.
- `ToolContext` gains `large_output_router` and `workshop_vars` fields plus
  the `with_large_output_router` builder; constructor defaults are `None` so
  sub-agents and test contexts are unaffected.
- `ToolRegistry::execute_full_with_context` applies routing after every tool
  call; `raw=true` in the tool input bypasses routing for that invocation.
- `EngineConfig` gains a `workshop` field; `Engine::new` creates the shared
  `WorkshopVariables` Arc when the field is present and wires it into every
  `build_tool_context` call.
- `Config` gains `[workshop]` table deserialization; `merge_config` propagates
  it like other optional tables.
- `config.example.toml` documents `[workshop]`, `large_output_threshold_tokens`
  (default 4096), and per-tool threshold overrides.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
wangfeng
2026-05-04 16:31:31 -07:00
parent 3cff070570
commit 5d1dee794d
10 changed files with 475 additions and 1 deletions
+20
View File
@@ -240,6 +240,26 @@ l3_threshold = 576000
cycle_threshold = 768000
seam_model = "deepseek-v4-flash"
# ─────────────────────────────────────────────────────────────────────────────────
# Workshop / Large-Output Routing (#548)
# ─────────────────────────────────────────────────────────────────────────────────
# Tool outputs exceeding `large_output_threshold_tokens` are routed through a
# V4-Flash synthesis sub-agent. Only the synthesis reaches the parent context;
# the raw text is stored in the workshop variable `last_tool_result` so the
# parent can call `promote_to_context` later if it needs the full content.
#
# Per-tool overrides let high-volume tools (e.g. exec_shell) use tighter
# thresholds without changing the global default.
#
# Add `raw = true` to any tool call to bypass routing for that invocation.
#
# [workshop]
# large_output_threshold_tokens = 4096
# [workshop.per_tool_thresholds]
# exec_shell = 2048 # shell output synthesised aggressively
# grep_files = 2048
# web_search = 8192 # web results can be large; give them more room
# ─────────────────────────────────────────────────────────────────────────────────
# Capacity Controller (runtime pressure guardrails)
# ─────────────────────────────────────────────────────────────────────────────────