Proposal: host-side AffixIO attestation alongside ADR observe/detect/prevent
I maintainer di solito rispondono entro 1 giorno
Nessuno ha ancora preso questa issue.
Valutazione
- Difficoltà
- 5/5
- Tempo stimato
- Più di una settimana
- Idoneità per principianti
- 25/100
Direzione di ricerca
Start by reading the README sections on Prevention and the Sensor schema; the issue does not identify specific files or tests. The first step is to get maintainer direction on whether this belongs in documentation, a Sensor schema change, or a local example, and whether to wait for Prevention OSS. Done would require an agreed scope and a minimal contribution matching it.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Descrizione
Summary
AffixIO can sit next to ADR as complementary host-side signed ACTION yes/no proof for consequential agent tool calls. ADR already observes activity, detects threats, and (in product) prevents unsafe actions. AffixIO would not replace that loop. It would add signed allow/deny evidence ADR can cite in detection/response, and optionally gate unsafe tool calls with host proof before execution.
Happy to stop at discussion, or contribute a minimal example if maintainers want one.
Why this fits
ADR Discovery/Sensor/Detection already cover inventory, telemetry, and threat detection. Prevention is called out in the README but is not in the current open-source release. Host-side attestation is a natural companion:
- ADR sees what the agent did or tried to do.
- AffixIO signs a concrete ACTION allow/deny decision on the host.
- Detection/response can treat that signature as citeable evidence, not another opaque policy black box.
Scope lock for AffixIO here: PII stays on the host. This is not person/age/KYC identity.
Product surface (AffixIO)
- Host-side signed ACTION yes/no via
npm i affixio - Local stdio MCP:
@affixio/[email protected] - Docs: https://www.affix-io.com/ and https://hub.affix-io.com/onboarding/
Proposed sit (no full wrap)
Optional complementary path only:
- Before a consequential tool call executes, request AffixIO host attestation for that ACTION.
- If denied, do not execute; emit signed deny evidence into the ADR observability/detection trail.
- If allowed, proceed and attach signed allow evidence so ADR Detection can cite it later.
This stays outside ADR core. No proposal to vendor AffixIO into Discovery/Sensor/Detection, and no full SDK wrap of ADR.
Ask for maintainers
- Does complementary host-side attestation belong in ADR's open-source direction (docs/example note, Sensor schema field for attestations, or wait until Prevention OSS lands)?
- If useful, would you prefer a short docs note plus a minimal local example under something like
Sensor/examples/, or hold until Prevention is public? - Any schema constraints for attaching signed evidence to Sensor events without bloating the unified telemetry model?
If maintainers want a minimal example, AffixIO can provide a small FIT-for-PR sketch (local stdio MCP + signed ACTION gate stub) and stop there. No large wrap planned.
- Lingua principale
- Python
- Stelle
- 2.1k
- Fork
- 230
- Merge medio
- 16h 8m
- PR unite (30g)
- 2
Preparare l'ambiente
- Nessun Dockerfile né file Docker Compose
- Ha un modello di pull request
- Leggi la guida per i contributori
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di uber/ADR
-
Difficoltà 5/5 Più di una settimana Idoneità per principianti 20/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 5/5 Più di una settimana Idoneità per principianti 32/100
I maintainer di solito rispondono entro 1 giorno
Issue simili
-
Input: placeholder drawn in the text colour when the empty Input is focused but not yet editedAperta
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
pierreaubert/gpui-toolkit#114 ·
I maintainer di solito rispondono entro 1 giorno
-
intake mcp-intake needs-human-review priority:high type:bug
Difficoltà 2/5 1-3 ore Idoneità per principianti 66/100
Ikalus1988/MisakaNet#4903 · 2 commenti ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 70/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 64/100
FinanceFlash/unvibecode#819 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 72/100
I maintainer di solito rispondono entro 1 giorno