obdash

justin/obdash

Fork 0

Commit Graph

Author	SHA1	Message	Date
justin	b5e0c96763	Fix #7 : derive action risk from UDS SIDs; fix response parsing Untrusted profiles could bypass the confirmation and responses were mis-parsed: - effective_risk(action): risk is now DERIVED from the actual service IDs the steps send — any write/actuator/reset/transfer SID (2F/31/11/14/2E/27/34-37/…) forces 'danger'; unknown SID / non-default session / security block force 'caution'. A profile can only RAISE risk, never label a reflash 'safe'. GUI gates the confirmation (and the risk badge) on this derived value. - Response checks use CONTIGUOUS subsequence matching + a hard '7F <sid>' negative-response guard, so an NRC data byte (e.g. 0x7E) can't false-pass as a positive response; applied to session/security/step checks. - 0x78 (responsePending) is treated as in-progress, not terminal failure. - controller.run_action restores slow ELM timing for the run (0x78 window). - Tests: risk-cannot-be-downgraded, NRC false-positive rejected, pending handled. Closes #7 Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016yT89n4zR4qbrySoSiEyZs	2026-07-01 19:29:44 -04:00
justin	d435384b58	#2 (framework): bi-directional / service-function engine Profile-defined UDS action sequences, run safely -- the framework for #2 (real per-vehicle actuator tests/resets are follow-on, added as verified profile data). - obdcore/actions.py: Action model + run_action() executing session (Mode 10) -> security (Mode 27 seed->key) -> command steps (2F/31/11/3E/... any hex) with positive/negative response checks. Security KEY algorithms are per-vehicle secrets and NOT bundled -- only trivial transforms (xor-ff/invert/add-ff) known; an action naming an unknown algorithm is BLOCKED (fails safe). Never synthesizes bytes -- runs only what the profile defines. validate_action() rejects malformed hex at load. - profile.py: load/save an actions[] block; ElmLink/MockLink read_raw(hex). - GUI: Diagnostics -> Service & Bi-directional dialog -- lists the profile's actions with risk badges; caution/danger gated behind a warning confirmation. - generic-obd2: two safe STANDARD actions (Tester-Present ping; ECU-Reset, caution + engine-off warning). PROFILE_SPEC.md documents the actions schema + safety rules. - tests/test_actions.py: runner, session+reset, security handshake, unknown-algo block, hex validation, profile load. All 5 suites pass. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016yT89n4zR4qbrySoSiEyZs	2026-07-01 16:33:51 -04:00

Author

SHA1

Message

Date

justin

b5e0c96763

Fix #7 : derive action risk from UDS SIDs; fix response parsing

Untrusted profiles could bypass the confirmation and responses were mis-parsed:

- effective_risk(action): risk is now DERIVED from the actual service IDs the
  steps send — any write/actuator/reset/transfer SID (2F/31/11/14/2E/27/34-37/…)
  forces 'danger'; unknown SID / non-default session / security block force
  'caution'. A profile can only RAISE risk, never label a reflash 'safe'. GUI
  gates the confirmation (and the risk badge) on this derived value.
- Response checks use CONTIGUOUS subsequence matching + a hard '7F <sid>'
  negative-response guard, so an NRC data byte (e.g. 0x7E) can't false-pass as a
  positive response; applied to session/security/step checks.
- 0x78 (responsePending) is treated as in-progress, not terminal failure.
- controller.run_action restores slow ELM timing for the run (0x78 window).
- Tests: risk-cannot-be-downgraded, NRC false-positive rejected, pending handled.

Closes #7

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016yT89n4zR4qbrySoSiEyZs

2026-07-01 19:29:44 -04:00

justin

d435384b58

#2 (framework): bi-directional / service-function engine

Profile-defined UDS action sequences, run safely -- the framework for #2 (real
per-vehicle actuator tests/resets are follow-on, added as verified profile data).

- obdcore/actions.py: Action model + run_action() executing session (Mode 10) ->
  security (Mode 27 seed->key) -> command steps (2F/31/11/3E/... any hex) with
  positive/negative response checks. Security KEY algorithms are per-vehicle
  secrets and NOT bundled -- only trivial transforms (xor-ff/invert/add-ff)
  known; an action naming an unknown algorithm is BLOCKED (fails safe). Never
  synthesizes bytes -- runs only what the profile defines. validate_action()
  rejects malformed hex at load.
- profile.py: load/save an actions[] block; ElmLink/MockLink read_raw(hex).
- GUI: Diagnostics -> Service & Bi-directional dialog -- lists the profile's
  actions with risk badges; caution/danger gated behind a warning confirmation.
- generic-obd2: two safe STANDARD actions (Tester-Present ping; ECU-Reset,
  caution + engine-off warning). PROFILE_SPEC.md documents the actions schema
  + safety rules.
- tests/test_actions.py: runner, session+reset, security handshake, unknown-algo
  block, hex validation, profile load. All 5 suites pass.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016yT89n4zR4qbrySoSiEyZs

2026-07-01 16:33:51 -04:00

2 Commits