Support binary HTTP response evidence

[tom-sapletta-com]

Adds one shared raw-byte classifier for classic OQL and Unified IR HTTP execution, with deterministic MIME, byte length, SHA-256 and magic evidence while preserving JSON/text compatibility.

Validation on the combined tree after PR #7:

• focused HTTP/interpreter/IR: 69 passed (original wider selection: 82 passed)
• full host suite: 1704 passed, 9 skipped
• E2E container: 1690 passed, 23 skipped
• production image build/start: TestQL 1.2.67
• Ruff, mypy, all Compose configs and governance: passed
• Viewer live integration: 59/59 passed

No PCB, SCH, Docker, Compose or Viewer source files are changed by this PR. Ready for protected validator review and merge.

[tom-sapletta-com]

Protected publication status: exact HEAD fe28201fd6d518da7aa38ae9e221904ffa26798e is mergeable and required CI test passed. Validator dispatch 33493557032 failed before repository validation at Create dependency Validator GitHub App token with GitHub API 401 A JSON web token could not be decoded. The failure is in validator credentials, not this PR; no new commit was pushed, so the frozen HEAD remains valid for retry after the App private key is restored.

[ifuri-validator-agent]

Deterministic Validator approval for exact head fe28201fd6d518da7aa38ae9e221904ffa26798e.

Ticket: ticket-002
Correlation ID: testql-pr-6-ticket-002
Model: zai/glm-5.3
Reviewed diff chunks: 3
Advisory LLM verdict: REQUEST_CHANGES
Advisory summary: Reviewed all 3 diff chunk(s). Governance/ticket documentation chunk for ticket-002 (binary HTTP response evidence). Manifest adds testql/** to core ownedPaths with updated lock hash; ticket scaffold is consistent. No executable source is present in this chunk; code review of the parser/executors must occur in other chunks. | Adds a shared bounded raw-byte HTTP body classifier (testql/http_response.py) used by the interpreter API runner, with deterministic evidence (kind, content-type, byte length, SHA-256, magic) and backwards-compatible JSON/text payload handling. Changes are confined to declared allowed paths, governance metadata is consistent, and tests pass. | Chunk refactors API response handling to use a shared parse_http_body that returns data plus deterministic evidence (kind, content_type, byte_length, sha256, magic) for JSON, text, SVG, and binary bodies in both the OQL interpreter and the IR executor. New tests cover classification, magic-byte precedence over content type, and evidence assertions in both runners. Tests pass.
Advisory findings: LLM verdict was REQUEST_CHANGES
The LLM output above is advisory and was not used as the approval trust root.
Merge will be attempted after this approval when explicitly authorized.

Decision record (recomputable)

DECISION D-002-2931
TICKET ticket-002
HEAD_SHA fe28201fd6d518da7aa38ae9e221904ffa26798e
CORRELATION_ID testql-pr-6-ticket-002
ACTOR agent:ifuri-validator-agent[bot]
APPLIED_RULE P-CORE-015
INPUT author_login = "tom-sapletta-com"
INPUT observed_checks = ["test=PASS"]
INPUT required_checks = ["test"]
INPUT required_checks_source = "protected registry (env/request)"
INPUT reviewer_login = "ifuri-validator-agent[bot]"
INPUT superseded_checks = []
VERDICT APPROVE AUTHORITY DETERMINISTIC
REJECTED REQUEST_CHANGES BECAUSE NO_UNSAFE_CHANGE_REASON_FOUND
ADVISORY llm_verdict = "REQUEST_CHANGES" MODEL "zai/glm-5.3"
ASSERT VERDICT_AUTHORITY != "ADVISORY"