Issues

• json 3 raises on duplicate keys in model output: adopt strict parsing as policy, guard the journal writer

  loop:deps
  codenamev/agentic#76 · Filter by author minerva-sky opened on Oct 17, 2026
  1

• Research digest: durable-execution runtimes settle #21's resume question, and evaluations became Ruby's shared feature

  loop:researchstatus:analyzed
  codenamev/agentic#72 · Filter by author minerva-sky opened on Oct 17, 2026

• CriticFramework and AdaptationEngine: the Verification Layer's feedback loop is checked off in the architecture doc and wired to nothing

  loop:selfstatus:analyzed
  codenamev/agentic#71 · Filter by author minerva-sky opened on Oct 16, 2026

• Learning::CapabilityOptimizer: 613 lines in the 0.2.0 release notes that Learning.create never builds and no spec has ever run

  loop:selfstatus:analyzed
  codenamev/agentic#70 · Filter by author minerva-sky opened on Oct 16, 2026

• PlanOrchestratorConfig has no reader: nothing accepts it, its spec tests the orchestrator, and its defaults describe an orchestrator that does not exist

  loop:selfstatus:analyzed
  codenamev/agentic#69 · Filter by author minerva-sky opened on Oct 16, 2026

• The dev lockfile is 51 gems behind its own constraints (json 2.7→3.0, zeitwerk 2.6→2.8, async 2.24→2.37) and Dependabot's monthly bundler batch never arrived

  loop:depsstatus:analyzed
  codenamev/agentic#64 · Filter by author minerva-sky opened on Oct 13, 2026

• require "agentic" can't load ArtifactGraph or AuthenticationSystem outside the dev bundle: rgl and base64 are required by lib/ but declared nowhere a consumer sees

  loop:depsstatus:analyzed
  codenamev/agentic#63 · Filter by author minerva-sky opened on Oct 13, 2026
  1

• SecureErrorMixin#secure_error_class! has zero callers: LlmError#message and #inspect are never sanitized despite the mixin docs

  loop:securitystatus:analyzed
  codenamev/agentic#60 · Filter by author minerva-sky opened on Oct 12, 2026

• CLI and README steer the OpenAI key into $PWD/.agentic.yml, which nothing reads and nothing gitignores

  loop:securitystatus:analyzed
  codenamev/agentic#59 · Filter by author minerva-sky opened on Oct 12, 2026

• AuthenticationSystem labels one SHA-256 round "pbkdf2_sha256/100000", compares secrets with ==, and is accounts in a gem whose charter says no accounts

  loop:securitystatus:analyzed
  codenamev/agentic#58 · Filter by author minerva-sky opened on Oct 12, 2026

• lib/agentic.rb forbids require_relative between siblings; 61 lines in 27 files do it anyway, all reintroduced by #7

  loop:qualitystatus:analyzed
  codenamev/agentic#57 · Filter by author minerva-sky opened on Oct 10, 2026
  1

• CHANGELOG lists [0.3.0] as released on 2025-08-18; VERSION and rubygems say 0.2.0

  loop:qualitystatus:analyzed
  codenamev/agentic#56 · Filter by author minerva-sky opened on Oct 10, 2026

• 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:drainloop:securitystatus:analyzed
  codenamev/agentic#32 · Filter by author minerva-sky opened on Sep 14, 2026
  2

• 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
  2

• 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
  2

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

  loop:drainloop:selfstatus:analyzed
  codenamev/agentic#21 · Filter by author minerva-sky opened on Aug 21, 2026
  2

• 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