Context
While using the integrating-jupiter skill (Portfolio playbook), /portfolio/v1/positions/{address} returned a 200 with:
elements: []
fetcherReports: []
tokenInfo: {}
This occurred consistently across multiple wallets, including the docs sample wallet.
Related docs issue: jup-ag/docs#828
Why this impacts the skill
Current guidance says to treat empty positions as valid state, which is generally correct. But when elements=[] and fetcherReports=[] for multiple addresses, this can indicate a suspicious backend/indexing condition rather than a truly empty portfolio.
In this state, an agent following the skill strictly may incorrectly conclude everything is fine.
Request
Please consider adding a Portfolio troubleshooting note in skills/integrating-jupiter/SKILL.md:
- Define an "empty-but-suspicious" condition (e.g.,
elements=[] + missing/empty fetcherReports).
- Recommend retries with backoff and
/portfolio/v1/platforms sanity check.
- Recommend checking status/known incidents and escalating with request metadata when persistent.
- Clarify that docs/sample wallet returning the same empty shape is a signal to escalate, not wallet-specific.
This would make agent behavior safer and reduce false "no positions" conclusions.
Context
While using the
integrating-jupiterskill (Portfolio playbook),/portfolio/v1/positions/{address}returned a200with:elements: []fetcherReports: []tokenInfo: {}This occurred consistently across multiple wallets, including the docs sample wallet.
Related docs issue: jup-ag/docs#828
Why this impacts the skill
Current guidance says to treat empty positions as valid state, which is generally correct. But when
elements=[]andfetcherReports=[]for multiple addresses, this can indicate a suspicious backend/indexing condition rather than a truly empty portfolio.In this state, an agent following the skill strictly may incorrectly conclude everything is fine.
Request
Please consider adding a Portfolio troubleshooting note in
skills/integrating-jupiter/SKILL.md:elements=[]+ missing/emptyfetcherReports)./portfolio/v1/platformssanity check.This would make agent behavior safer and reduce false "no positions" conclusions.