-
-
Notifications
You must be signed in to change notification settings - Fork 2
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
[FIX] PR #1550 CI failures — _is_unattended_context too aggressive in test env
bugSomething isn't workingSomething isn't workingfixBug or fixBug or fixintrospectionFound by session introspectionFound by session introspectionStatus: Open.#1554 In Lexus2016/hermes-agent-evolution;[FIX] Suppress terminal retries on pending_approval — 14/32 failures (44%) are retry spirals
acceptedAccepted by evolution — sent to a PR / implementedAccepted by evolution — sent to a PR / implementedbugSomething isn't workingSomething isn't workingintrospectionFound by session introspectionFound by session introspectionStatus: Open.#1542 In Lexus2016/hermes-agent-evolution;[UPSTREAM-ESCALATION] Sync release v2026.7.30 — 3087 commits, 85 conflicts (16 in core runtime)
escalationRequires owner judgementRequires owner judgementupstream-syncUpstream sync tasksUpstream sync tasksStatus: Open.#1526 In Lexus2016/hermes-agent-evolution;[Eval Harness] Step 3-4: Null-agent baseline + run baseline, output scores.jsonl
enhancementNew feature or requestNew feature or requestneeds-splitWanted, but exceeds one cycle — needs decompositionWanted, but exceeds one cycle — needs decompositionStatus: Open.#1516 In Lexus2016/hermes-agent-evolution;[Eval Harness] Step 2: Agent runner — execute task, capture full trajectory
enhancementNew feature or requestNew feature or requestneeds-splitWanted, but exceeds one cycle — needs decompositionWanted, but exceeds one cycle — needs decompositionStatus: Open.#1515 In Lexus2016/hermes-agent-evolution;[MCP Stateless] Slice B-3: Mock stateless server + integration tests for the stateless path
enhancementNew feature or requestNew feature or requestmcpMCP adapter / protocolMCP adapter / protocolneeds-splitWanted, but exceeds one cycle — needs decompositionWanted, but exceeds one cycle — needs decompositionresearch-drivenCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1513 In Lexus2016/hermes-agent-evolution;[MCP Stateless] Slice B-2: Stateless request routing — skip initialize, add _meta inline + server/discover
enhancementNew feature or requestNew feature or requestmcpMCP adapter / protocolMCP adapter / protocolneeds-splitWanted, but exceeds one cycle — needs decompositionWanted, but exceeds one cycle — needs decompositionresearch-drivenCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1512 In Lexus2016/hermes-agent-evolution;[MCP Stateless] Slice B-1: Capability-detection flag + synthesize_capabilities()
acceptedAccepted by evolution — sent to a PR / implementedAccepted by evolution — sent to a PR / implementedenhancementNew feature or requestNew feature or requestmcpMCP adapter / protocolMCP adapter / protocolresearch-drivenCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1511 In Lexus2016/hermes-agent-evolution;[IMPROVEMENT] Harness Control System hardening: adopt cold-start trap, crash-isolation, and cache-measurement pitfalls from 729-config harness study (static beats online at practical budgets)
enhancementNew feature or requestNew feature or requestproposalEvolution-generated improvement proposalEvolution-generated improvement proposalresearch-generatedCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1509 In Lexus2016/hermes-agent-evolution;[FEATURE] RSIBench-Data: meta-evaluation of pipeline research stage — checkpoint preservation rule + four trajectory diagnostic patterns (78% regression-after-peak signal)
enhancementNew feature or requestNew feature or requestproposalEvolution-generated improvement proposalEvolution-generated improvement proposalresearch-generatedCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1508 In Lexus2016/hermes-agent-evolution;[IMPROVEMENT] Earned Authority: authorization continuity for evolving agents — state-bound model with transition envelope + immutable effect ceiling
enhancementNew feature or requestNew feature or requestproposalEvolution-generated improvement proposalEvolution-generated improvement proposalresearch-generatedCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1507 In Lexus2016/hermes-agent-evolution;[IMPROVEMENT] WML: Node-Mechanism Attribution for structured skill debugging — identify failed workflow node, implicated mechanisms, smallest valid edit target
enhancementNew feature or requestNew feature or requestproposalEvolution-generated improvement proposalEvolution-generated improvement proposalresearch-generatedCreated by the evolution research cycleCreated by the evolution research cycleStatus: Open.#1506 In Lexus2016/hermes-agent-evolution;