Commit graph

5 commits

Author SHA1 Message Date
aac4d1f2ce chore: pin chain-module-sdk v0.3.0 and switch to #[chain_module]
The pre-rename #[fai_module] attribute is gone in SDK 0.3.0; the
dependency is tag-pinned so an SDK push can no longer change what
this module builds against. Rebuilt artifact where tracked.

Signed-off-by: flemming-it <sf@flemming.it>
2026-07-12 13:52:37 +02:00
a0e713d752 fix: redact grouped German IBAN (DE89 3704 …) as one IBAN, not partial+phone
Signed-off-by: flemming-it <sf@flemming.it>
2026-07-07 20:22:12 +02:00
ffc9de6bc3 feat: detect grouped German tax-id (2-3-3-3) + honest coverage/warning in report
The GERMAN_TAX_ID category now matches the official spaced form
'12 345 678 901' on top of the contiguous 11-digit form. The JSON
report gains a coverage object declaring what is NOT detected
(names list-only, no addresses) and a warning that the report
itself embeds the original PII.

Signed-off-by: flemming-it <sf@flemming.it>
2026-07-07 19:18:56 +02:00
3485a51941 refactor: depend on renamed chain-module-sdk
Signed-off-by: flemming-it <sf@flemming.it>
2026-06-16 11:28:57 +02:00
734d8f6e6f feat: text.anonymize v0.1.0 — regex-based PII redaction
First release. Pure-Rust, in-WASM, regex-only, declares no
permissions. Suitable as a first redaction pass after
text.extract before any cloud-LLM step.

Detection categories:

  EMAIL          RFC-5321-ish local@domain, IDN-aware.
  PHONE          International (+CC …) and DE national
                 (030 …, 0151-…) shapes, 7..20 raw digits.
  IBAN           Word-bounded [A-Z]{2}\d{2}[A-Z0-9]{11,30}.
                 Structural only — MOD-97 checksum
                 deliberately skipped so partial / truncated
                 tokens in running text still get redacted.
  BIC            8 or 11 uppercase alnum.
  IPV4           Four 0..255 octets, dot-separated.
  GERMAN_TAX_ID  11 consecutive digits, word-bounded.
  CUSTOM         Operator-supplied bare terms from the
                 newline-separated `custom_terms` input,
                 matched whole-word case-insensitive.

Token shape: ⟦TYPE_N⟧ — U+27E6 / U+27E7 mathematical white
square brackets. Distinct from any plain ASCII `[…]` already
present in source text (Markdown links, legal citations,
code blocks) so a reviewer never has to guess which `[…]`
is a redaction.

Outputs:

  anonymized  text  Input with PII replaced by ⟦TYPE_N⟧.
                    Counter restarts at 1 per type so the
                    tokens stay operator-readable.
  report      json  { redactions: [{type, token, original,
                    offset}…], counts: { TYPE: n, … } }.
                    Full original-text reconstruction is
                    possible from this — the GDPR
                    Art. 32(1)(a) "ability to undo"
                    requirement.

Quality bar (7 unit tests):
  * email round-trip
  * IBAN + BIC don't eat each other
  * three phone-number shapes redact
  * IPv4 only matches valid 0..255 octets
  * custom_terms case-insensitive
  * no double-redaction on overlapping patterns
  * per-category counter resets correctly

Built artefact: target/wasm32-wasip2/release/text_anonymize.wasm
(~180 KiB stripped).

NER for free-text names / organisations / locations is the
v0.2.0 plan once a benchmarked ONNX model is selected; the
operator's `custom_terms` field is the v0.1.0 escape hatch.

Signed-off-by: flemming-it <sf@flemming.it>
2026-05-25 20:53:24 +02:00