Add R4 escalation reminders and the structured source-id evidence ledger (E4) #131

Merged
weichen merged 1 commits from feature/agent-e4-escalation-ledger into main 2026-10-10 17:50:21 +00:00
Owner

E4-MiniMax 剩余项(提案 §3 R4 + E3 账本结构化):①R4 升级提醒——连续 2 次工具失败/检索连续空结果时 harness 注入换路指令(成败计数在 onEvent 终点单一事实源;每条件每链一次,仿双速提醒的「具体指令」经验);②结构化 source_id 账本——证据类工具(knowledge/web/tavily_search、fs_read_file、工作区 read_file/grep)成功输出登记 src_N 并尾注【src_N|工具】;事实核查提示指向来源;③base.deliverables v7 + agent-sub-research v3 说明标记。

Evidence: go test ./... 全绿 + 4 个新回归锚点;eval 前后对照在 Dev 侧(裁定 6)。

Refs: Dev 任务 feature/agent-e4-escalation-ledger。

E4-MiniMax 剩余项(提案 §3 R4 + E3 账本结构化):①R4 升级提醒——连续 2 次工具失败/检索连续空结果时 harness 注入换路指令(成败计数在 onEvent 终点单一事实源;每条件每链一次,仿双速提醒的「具体指令」经验);②结构化 source_id 账本——证据类工具(knowledge/web/tavily_search、fs_read_file、工作区 read_file/grep)成功输出登记 src_N 并尾注【src_N|工具】;事实核查提示指向来源;③base.deliverables v7 + agent-sub-research v3 说明标记。 Evidence: go test ./... 全绿 + 4 个新回归锚点;eval 前后对照在 Dev 侧(裁定 6)。 Refs: Dev 任务 feature/agent-e4-escalation-ledger。
weichen added 1 commit 2026-10-10 17:41:51 +00:00
Why: The plan's E4 remainder under the MiniMax-only constraint: the
harness must detect the model going in circles and inject a
change-of-strategy instruction (R4 - the model cannot be trusted to
notice its own repeat-fail loop), and number sourcing needs addressable
evidence: tool results carry [src_N|tool] tags so "found in materials"
becomes "found in source src_N" for both the model and the fact-check
notice.

Why: Failure counting lives in one place (onEvent end, keyed by the
slot status) because tool execution is parallel; the ledger and tag
append are serialized by the same mutex, and escalation reminders fire
once per condition streak - following the dual-speed nudge finding that
MiniMax ignores vague reminders but follows concrete instructions.

Evidence: go test ./... green; new anchors TestEscalationOnConsecutive
ToolFailures, TestEscalationOnZeroHitSearches, TestEvidenceLedger
TagsSource, TestEvidenceLedgerSkipsNonEvidenceTools. Live before/
after comparison runs on the Dev side per ruling 6.

Impact: knowledge/web/tavily searches, fs_read_file and the workspace
read_file/grep results now end with a source tag and enter the run
ledger (2k-char head); two consecutive tool failures or two empty
search results inject a concrete change-of-strategy reminder;
base.deliverables v7 and agent-sub-research v3 explain the tags;
fact-check notice points at src results. Non-evidence tools and
failures are unchanged.

Refs: proposal 2026-10-10-agent-delegation-routing-sandbox.md (E4,
R4); Dev task feature/agent-e4-escalation-ledger.
weichen merged commit d116eda0bc into main 2026-10-10 17:50:21 +00:00
Sign in to join this conversation.
No Reviewers
No labels
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: weichen/Agent-System-Backend#131