Issues

• Switch skill-env-reachability-test.sh from hand-maintained SANDBOX_RUNTIME_VARS to the released-tag sandbox env manifest

  blocked
  fullsend-ai/agents#1738 · Filter by author jflowers opened on Oct 21, 2026
  2

• eval: 965 leftover eval-* repos on the halfsend account, and a one-off cleanup

  pr-open
  fullsend-ai/agents#1736 · Filter by author jwm4 opened on Oct 21, 2026
  2

• Feature: Fullsend CI Failure Root-Cause Analysis, Learning, and HITL Jira Escalation

  fullsend-ai/agents#1735 · Filter by author rrasouli opened on Oct 21, 2026

• Retro and Triage ignore issue-target permissions declared in config presets

  bugready-to-code
  fullsend-ai/agents#1728 · Filter by author ederign opened on Oct 21, 2026
  3

• post-triage: re-categorization leaves stale category label on the issue

  bugready-to-code
  fullsend-ai/agents#1725 · Filter by author rh-hemartin opened on Oct 21, 2026
  3

• pre-review: dismiss the review bot's stale approval at review run start (follow-up to #1719)

  bugready-to-code
  fullsend-ai/agents#1720 · Filter by author waynesun09 opened on Oct 20, 2026
  3

• Filter sub-agent infrastructure failures from fix-agent inputs and skip dispatch on non-code findings

  fullsend-ai/agents#1714 · Filter by author fullsend-ai-retro opened on Oct 20, 2026

• fix-agent: discover trusted repository skills on demand

  featureready-for-triagetriaged
  fullsend-ai/agents#1713 · Filter by author ifireball opened on Oct 20, 2026
  3

  

• eval: bump agent-eval-harness to v1.56.3 and harden the shared eval scripts

  pr-open
  fullsend-ai/agents#1712 · Filter by author jwm4 opened on Oct 19, 2026
  3

• Retro agent should verify and attach repo evidence when filing issues derived from third-party tool comments

  fullsend-ai/agents#1711 · Filter by author fullsend-ai-retro opened on Oct 19, 2026

• review: keep dismissed_by_human when only the category changes at the same file:line

  bugready-to-code
  fullsend-ai/agents#1709 · Filter by author Benkapner opened on Oct 19, 2026
  4

  

• post-review.sh: pass the review-findings-v2 projection to fullsend post-review via a trusted input instead of the review body

  fullsend-ai/agents#1706 · Filter by author fullsend-ai-triage opened on Oct 19, 2026

• Support running selected functional-eval cases locally

  help wantedneeds-info
  fullsend-ai/agents#1704 · Filter by author Benkapner opened on Oct 18, 2026
  4

  

• eval: distinguish PR discovery failures from an empty PR list

  bugready-to-code
  fullsend-ai/agents#1702 · Filter by author Benkapner opened on Oct 18, 2026
  4

  

• eval: preserve first-attempt artifacts when retrying failed cases

  triaged
  fullsend-ai/agents#1701 · Filter by author Benkapner opened on Oct 18, 2026
  4

  

• intent-coherence: cap AGENTS.md rule-derived findings at the action the rule prescribes; don't invent compliance requirements; consult the cited ADR's migration phase

  fullsend-ai/agents#1692 · Filter by author fullsend-ai-retro opened on Oct 15, 2026

• CONTRIBUTING.md / PR template: require a linked, agreed-upon issue before PRs that add new controls or touch protected paths

  documentationready-for-triageready-to-code
  fullsend-ai/agents#1684 · Filter by author fullsend-ai-retro opened on Oct 15, 2026
  3

• pr-review: merge protected-path and missing-authorization findings when both stem from the same missing linked issue

  bugready-for-triageready-to-code
  fullsend-ai/agents#1683 · Filter by author fullsend-ai-retro opened on Oct 15, 2026
  3

• post-review/post-triage control-label guard is case-sensitive; agent-recommended removes can strip pipeline labels like requires-manual-review

  ready-for-triagetriaged
  fullsend-ai/agents#1682 · Filter by author fullsend-ai-retro opened on Oct 15, 2026
  2

• github-forge skill GraphQL recipes use invalid -F number:= syntax, breaking linked-PR and sub-issue lookups

  bugready-for-triageready-to-code
  fullsend-ai/agents#1681 · Filter by author fullsend-ai-retro opened on Oct 15, 2026
  3

• Proposal: maintainer-controlled fix-loop budgets per PR

  featuretriaged
  fullsend-ai/agents#1680 · Filter by author Benkapner opened on Oct 15, 2026
  4

• Harden secret scanning of agent-authored issue context and PR descriptions

  triaged
  fullsend-ai/agents#1673 · Filter by author shairevivo opened on Oct 13, 2026
  2

• pr-risk-assessment skill path is not runtime-neutral

  fullsend-ai/agents#1672 · Filter by author jflowers opened on Oct 13, 2026
  2

• Add intra-run multi-pass convergence loop to review orchestrator

  enhancementfeaturetriaged
  fullsend-ai/agents#1670 · Filter by author jflowers opened on Oct 13, 2026
  5

• Fix agent: conflict resolution should reconcile behavioural overlap with incoming main changes, not defer it as out of scope

  fullsend-ai/agents#1666 · Filter by author fullsend-ai-retro opened on Oct 12, 2026