mirror of
https://github.com/bytedance/deer-flow.git
synced 2026-11-20 11:20:35 +00:00
* feat(frontend): improve file attachment recovery and artifact presentation * feat(extensions): add Office artifact previews with search and navigation * fix: preserve public viewers and HTTP Office previews * fix: address file preview and upload review findings * fix: use typed upload errors and linear bookmark allocation * fix(office-preview): isolate parsing and bound preview resources * fix(ci): install Office preview fixture dependencies for E2E * fix(ci): provision matching Playwright browsers before E2E tests * fix(ci): isolate Python libraries and include Office documentation * test(frontend): wait for download status to finish rendering * fix(harness): account for tool results in loop detection * fix(harness): make tool frequency detection observational * fix(harness): support blocked and waiting plan todos * style(harness): wrap overlong system prompt lines * Improve loop detection and todo message formatting Refactor middleware logic for handling repeated results and update messages for todo list. Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com> * Address plan progress and loop detection review --------- Co-authored-by: Willem Jiang <willem.jiang@gmail.com> Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
39 lines
2.0 KiB
Markdown
39 lines
2.0 KiB
Markdown
# Loop detection lifecycle
|
|
|
|
`LoopDetectionMiddleware` owns call-pattern detection. Its place in the chain
|
|
and audit contract are documented in the
|
|
[middleware guide](../packages/harness/deerflow/agents/middlewares/AGENTS.md).
|
|
|
|
## Run-scoped state
|
|
|
|
Hash histories, frequency windows/counters, and frequency-metric reporting
|
|
sets share a `(thread_id, run_id)` key. This gives a fresh budget to each user
|
|
run when a compiled graph is cached and reused, while keeping one budget
|
|
across repeated graph entries belonging to the same Gateway run (including
|
|
hidden goal continuations). `after_agent` clears only transient pending
|
|
warnings for its own scope, not those histories; the sync/async
|
|
`before_agent` hooks remain topology-preserving no-ops and must not delete a
|
|
sibling run's pending warning. Direct LangGraph embedders may omit
|
|
`context.run_id`; that fallback is anchored to the invocation's shared
|
|
`Runtime.control` object and mapped to an opaque generated ID, because
|
|
LangGraph replaces `Runtime` per node and CPython can reuse freed object
|
|
addresses. `after_agent` releases the anchor mapping, while the bounded map
|
|
covers abnormal exits. The compatibility-named
|
|
`max_tracked_threads` limit bounds run scopes, and `reset(thread_id)` clears
|
|
every retained run scope for that thread.
|
|
|
|
## Decision ordering
|
|
|
|
Hash-based loop decisions are evaluated before the frequency metric is updated.
|
|
A hash warning does not prevent the remaining tool names in the same response
|
|
from being counted. Frequency thresholds only emit INFO metrics with
|
|
`action="observe_only"`; they never enqueue a model warning, strip tool calls,
|
|
or set a stop reason. The hash warning and hard-stop behavior remains scoped to
|
|
identical calls whose results show no progress.
|
|
|
|
The lead agent applies the subagent call limit before loop detection, so the
|
|
detector reconciles results against the admitted batch. A successful
|
|
`write_file` or `str_replace` invalidates the predicted result for older
|
|
reads of the same path. The next read observes the current version and
|
|
completed results remain in the repetition window.
|