Issues

• Planner never sees the capability catalog: plan-time capability selection is the missing half of #20

  loop:researchstatus:analyzed
  codenamev/agentic#49 · Filter by author minerva-sky opened on Sep 20, 2026

• Token usage is parsed on every LLM response and dropped before the journal, TaskResult, or the learning hooks see it

  loop:selfstatus:analyzed
  codenamev/agentic#46 · Filter by author minerva-sky opened on Sep 19, 2026
  2

• Saved agents silently lose capabilities on any capability version bump; compatible_with? has zero callers

  loop:selfstatus:analyzed
  codenamev/agentic#45 · Filter by author minerva-sky opened on Sep 19, 2026
  1

• Self-assembly can't see the registry: keyword inference is a hardcoded 11-name list, and only 4 of 7 registered capabilities are on it

  loop:selfstatus:analyzed
  codenamev/agentic#44 · Filter by author minerva-sky opened on Sep 19, 2026

• CI tests async 2.24 while consumers on maintained Rubies resolve 2.46: the 3.2 caps in Gemfile freeze every matrix leg

  loop:depsstatus:analyzedstatus:blocked
  codenamev/agentic#42 · Filter by author minerva-sky opened on Sep 16, 2026

• PII sanitizer rewrites UUIDs: ~10% of task IDs in TaskFailure records come back as [REDACTED_PHONE]

  loop:depsstatus:analyzed
  codenamev/agentic#41 · Filter by author minerva-sky opened on Sep 16, 2026

• Sanitizer rejects any Markdown artifact with inline code as command injection

  loop:securitystatus:analyzed
  codenamev/agentic#32 · Filter by author minerva-sky opened on Sep 14, 2026

• SecurityError from Workspace#add_artifact bypasses every rescue between the LLM and the plan record

  loop:securitystatus:analyzed
  codenamev/agentic#31 · Filter by author minerva-sky opened on Sep 14, 2026
  1

• Research digest: the plan-as-data bet is uncontested in Ruby, and roast just vacated the lane

  loop:researchstatus:analyzed
  codenamev/agentic#23 · Filter by author minerva-sky opened on Aug 23, 2026
  1

• Provider strategy: the researched proposal WORLD.md asks for (measured coupling, three options, one lean)

  loop:selfstatus:analyzed
  codenamev/agentic#22 · Filter by author minerva-sky opened on Aug 21, 2026
  1

• The durable journal can replay but nothing can resume: crash recovery stops one API call short of the payoff

  loop:selfstatus:analyzed
  codenamev/agentic#21 · Filter by author minerva-sky opened on Aug 21, 2026

• Planned plans are flat: the LLM planner can't express the dependency graph the orchestrator already executes

  loop:selfstatus:analyzed
  codenamev/agentic#20 · Filter by author minerva-sky opened on Aug 21, 2026

• Six runtime dependencies have no version constraint, including ruby-openai (locked at 7.1.0, consumers resolve 8.3.0)

  loop:depsstatus:analyzed
  codenamev/agentic#16 · Filter by author minerva-sky opened on Aug 19, 2026

• Ruby version currency: the gemspec claims >= 3.0.0, CI tests only 3.2.4, and all three of those Rubies are EOL

  loop:depsstatus:analyzed
  codenamev/agentic#15 · Filter by author minerva-sky opened on Aug 19, 2026

• Harden PersistentAgentStore path handling and capability-filter dispatch against untrusted agent configs

  loop:securitystatus:analyzed
  codenamev/agentic#12 · Filter by author minerva-sky opened on Aug 17, 2026
  1