Universal video downloader
Primary reason: platform-adversarial maintenance risk.
Demand exists, but the category tends to become a fragile arms race against platform rules and creates policy/trust risk that does not fit EXFILTR.
// rejected ideas
EXFILTR rejects more ideas than it builds.
Some ideas have real demand but fail the trust, maintenance, permission, monetization, distribution, or browser-native filters.
// Archive
Public-safe rejection notes.
These are not permanent moral judgments. They are current operating decisions based on EXFILTR's trust and maintenance rules.
Generic AI summarizer extension
Primary reason: weak workflow ownership.
Generic summarization is easy to copy, hard to trust, and too broad. Refine stays focused on browser-native readability rather than becoming AI feature soup.
Pushbullet-style relay
Primary reason: infrastructure and support burden.
Cross-device relay workflows may monetize, but they require backend reliability, notification handling, and support complexity that should wait.
Broad new-tab productivity dashboard
Primary reason: commodity and bloat risk.
The category is crowded and often becomes dashboard-heavy. EXFILTR prefers focused workflow infrastructure with clearer retention and trust advantage.
Archive entries are not launch claims. A rejected, deferred, or interesting idea is not live, not approved, and not a product commitment.
// Rejection filters
Why ideas get killed.
EXFILTR can reject an idea even when there is demand. Trust, maintenance, and retention matter more than novelty.
Too much access.
If the workflow requires broad, hard-to-explain browser access, it is rejected or narrowed until trust is defensible.
Adversarial category.
Products that depend on fighting platform policies, fragile scraping, or constant breakage are not launch candidates.
Too heavy to own.
High-support, high-infrastructure, or reliability-heavy products wait until EXFILTR can maintain them calmly.
Not a habit.
Low-frequency novelty tools are rejected unless the workflow has clear repeated utility and day-three retention potential.
No workflow ownership.
Utilities that could be copied as a generic feature or prompt wrapper do not create the trust moat EXFILTR needs.
AI is not the product.
AI can support research or workflows, but generic AI layers without a browser-native job are rejected.