Issues

• bug: findings.schema.json rejects scan.remit_version that v2 emits and schema.py accepts

  bug
  open-agent-ai-security/praxen#266 · Filter by author virtualsteve-exa opened on Oct 9, 2026

• Add praxen to awesome-ai-plugins?

  open-agent-ai-security/praxen#265 · Filter by author zerocodefast opened on Sep 22, 2026

• docs site: mermaid diagrams render before the web font loads, so every node label is clipped by a glyph

  open-agent-ai-security/praxen#264 · Filter by author virtualsteve-exa opened on Sep 10, 2026

• 2.0 GA: switch the production catalog install to the vendored payload (same route as beta)

  open-agent-ai-security/praxen#263 · Filter by author virtualsteve-exa opened on Sep 1, 2026

• x-high: adjudicator brief requires raw-run evidence checkpoints but never hands over their paths

  open-agent-ai-security/praxen#261 · Filter by author virtualsteve-exa opened on Aug 20, 2026
  2

• v1.3-opus5 uagents baseline: finding -007 tags ASI02 on a missing approval gate — KB arbitration says 'never ASI02 for the gate itself' (1.4 / re-baseline)

  bug
  open-agent-ai-security/praxen#257 · Filter by author virtualsteve-exa opened on Aug 17, 2026

• v1.3-opus5 finbot baseline: unauthenticated /api/users CRUD (user.py) is uncovered — outside -001's admin scope, no other finding files it (1.4 / re-baseline)

  bug
  open-agent-ai-security/praxen#256 · Filter by author virtualsteve-exa opened on Aug 17, 2026

• Scanner misses client-side call sites that target undeclared routes — finbot's operator approve/reject buttons 404 while the unauthenticated path works (1.4 / re-baseline)

  bug
  open-agent-ai-security/praxen#255 · Filter by author virtualsteve-exa opened on Aug 17, 2026

• v1.3-opus5 uagents baseline: finding -013 carries ASI08 against the KB's own arbitration — pure alerting gap is RAISE-only (1.4 / re-baseline)

  bug
  open-agent-ai-security/praxen#254 · Filter by author virtualsteve-exa opened on Aug 16, 2026

• v1.3-opus5 uagents baseline: finding -002 overstates the Helm Secret wiring — the chart renders the Secret but nothing delivers it to the pod (1.4 / re-baseline)

  bug
  open-agent-ai-security/praxen#253 · Filter by author virtualsteve-exa opened on Aug 16, 2026

• Scanner emits inconsistent evidence-path prefixes on multi-root workspaces — inflates unique findings in x-high (1.4 / re-baseline)

  open-agent-ai-security/praxen#252 · Filter by author virtualsteve-exa opened on Aug 13, 2026

• Dominant-path ladder applies to Monitor Continuously, which has no data path to rank — fails the rule's own scoping test (score-moving — 1.4 / re-baseline)

  open-agent-ai-security/praxen#251 · Filter by author virtualsteve-exa opened on Aug 13, 2026

• Step 8b M5 can match Praxen's own ./reports/ output — the scanner credits the target for having been scanned (score-moving — 1.4 / re-baseline)

  open-agent-ai-security/praxen#250 · Filter by author virtualsteve-exa opened on Aug 13, 2026

• Step 9.4 scores from positives[] but positives are never persisted — the one input of four that can't survive a compaction (score-moving — 1.4 / re-baseline)

  open-agent-ai-security/praxen#249 · Filter by author virtualsteve-exa opened on Aug 13, 2026

• Step 8b M1 misses security tests named after the control they guard (score-moving — 1.4 / re-baseline)

  enhancement
  open-agent-ai-security/praxen#248 · Filter by author virtualsteve-exa opened on Aug 13, 2026

• High-mode remit feedback reaches the operator only via ephemeral chat — durable report can cite a rule the audit found defective

  enhancement
  open-agent-ai-security/praxen#247 · Filter by author virtualsteve-exa opened on Aug 12, 2026

• Bug: render.py redaction backstop over-redacts ${VAR} references as secret values

  open-agent-ai-security/praxen#243 · Filter by author virtualsteve-exa opened on Aug 11, 2026
  1

• Docs review (consistency): schema version documented as 2.0 (is 3.0), SECURITY.md still says pre-1.0, + 6 more fix-soon

  open-agent-ai-security/praxen#226 · Filter by author virtualsteve-exa opened on Aug 3, 2026
  2

• Manifest authoring is the pipeline's fragile step: large exact-format document written entirely by the model, no tooling assist

  open-agent-ai-security/praxen#217 · Filter by author virtualsteve-exa opened on Aug 1, 2026

• RFE: remit over-reach cleanup — 5 remaining v1.2 targets (helperbot, craftbot, salesforce, autogen, uagents) — defer + re-baseline later

  open-agent-ai-security/praxen#201 · Filter by author virtualsteve-exa opened on Jul 29, 2026
  1

• RFE: aider remit cleanup — fabricated allowlist, header-as-rule, documented-feature over-reach (defer + re-baseline later)

  enhancement
  open-agent-ai-security/praxen#200 · Filter by author virtualsteve-exa opened on Jul 29, 2026
  2

• RFE: Remit generator should author well-formed, docs-grounded, non-over-reaching statements (heading-as-rule + fabrication/over-scope)

  enhancement
  open-agent-ai-security/praxen#198 · Filter by author virtualsteve-exa opened on Jul 29, 2026

• RFE: Thinking Modes — opt-in high-fidelity accuracy tiers for users (High: post-scan FP review; X-High: 3-scan consensus adjudication)

  enhancement
  open-agent-ai-security/praxen#197 · Filter by author virtualsteve-exa opened on Jul 29, 2026
  2

• RFE: Tighten Step 8.5 fold-vs-break-out decomposition rule to reduce finding-count variance

  enhancement
  open-agent-ai-security/praxen#196 · Filter by author virtualsteve-exa opened on Jul 29, 2026
  1

• RFE: Sharpen RAISE category band anchors to reduce band-edge scoring variance on mid-maturity targets

  enhancement
  open-agent-ai-security/praxen#195 · Filter by author virtualsteve-exa opened on Jul 29, 2026
  1