phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-27 23:22:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

add context to distinguish multiple credentials for the same (user, provider)

Browse source
This commit is contained in:
Ashwin Bharambe 2025-05-20 19:41:43 -07:00
parent b43cdaaed5
commit f6fb4a5865
3 changed files with 23 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
llama_stack/apis/credentials/credentials.py
View file

@ -59,8 +59,9 @@ class Credentials(Protocol):
token: str,
nonce: str | None = None,
ttl_seconds: int = 3600,
context: dict[str, str] | None = None,
) -> str:
"""Create a new set of credentials for a given provider.
"""Create a new set of credentials for a given provider. The unique key is (provider_id, context)
:param provider_id: The ID of the provider to create credentials for.
:param token_type: The type of token to create. This is provided in the API to serve as lightweight
@ -70,6 +71,9 @@ class Credentials(Protocol):
:param ttl_seconds: The time to live for the credential in seconds. Defaults to 3600 seconds.
When token_type is oauth2_authorization_code, the TTL is ignored and is obtained
from the provider when exchanging the code for an access token.
:param context: Context to distinguish between different credentials for the same provider. For example,
for the MCP provider, this could be the toolgroup_id (which corresponds to a unique MCP
endpoint).
:returns: The ID of the created credential.
"""
...