►
From YouTube: Tool House Rock #13 | Llama
Description
MakerDAO Forum Post: https://forum.makerdao.com/t/tool-house-rock-13-llama-tuesday-august-8th-2023-14-00-utc/21540/
Description:
Tool House Rock is a structured community call to give projects a way to showcase their tools and solutions to the growing MakerDAO Ecosystem. Tool House Rock will follow a consistent call agenda, focused on giving Ecosystem Actors a foundation for reviewing and assessing tools.
This Session's Guests:
* Llama - Shreyas Hariharan - https://twitter.com/HelloShreyas
What is Llama?
Llama 6 is an onchain governance framework for protocols. It enables protocols to encode access control policies, create and execute actions, and manage funds programmatically. Llama helps protocols run their operations with 10 members or 10,000. Through Llama, functions can be permissioned granularly and access control policies can be issued to individuals, multisigs, or token voting DAO contracts.
Llama is designed to:
Encode access control policies
Define action execution rules
Create and execute actions
Manage funds programmatically
Progressively decentralize
Improve the transparency and auditability of decision making
Disclaimer
This communication is provided for information purposes only. The views and opinions expressed in this meeting are those of people involved and do not reflect the official policy or position of MakerDAO or any of its contributors, Core Units, or affiliates. This communication makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only and should not be relied upon as legal, business, investment, financial, or tax advice. You should consult your own advisers as to those matters. References to any digital assets and the use of finance-related terminology are for illustrative purposes only and do not constitute any recommendation for any action or an offer to provide investment, financial, or other advisory services.
MakerDAO Links:
https://makerendgame.com/toolshed
https://mips.makerdao.com
https://forum.makerdao.com
https://chat.makerdao.com
A
A
A
few
disclaimers
before
we
get
started,
our
views
are
our
own
and
are
not
representative
of
maker
as
a
whole.
Inaccuracies
are
possible.
Information
moves
very
quickly
in
the
space
and
especially
if
you're
viewing
this
in
a
distant
future,
it's
possible
that
things
have
changed,
and
finally,
none
of
this
is
legal
or
financial
advice.
B
Awesome
thanks
for
having
me
and
definitely
excited
to
be
here,
I
think
when
I
first
I
got
really.
You
know
deeper
into
ethereum
and
epari,
because
system
Miko
is
definitely
one
of
the
first
projects
that
I
found
interesting,
and
you
know
it
was
one
of
the
first
sort
of
decentralized
collectives
that
were
on
Shane
and
that
was
quite
inspiring
to
me
and-
and
you
know
one
of
one,
among
other
reasons,
that
I
got
kind
of
more
involved
in
in
protocols
and
endow
governance,
so
definitely
excited
to
be
here.
B
So,
taking
a
step
back,
it's
the
evolution
of
on-chain
governance.
We
you
know.
First,
a
lot
of
teams
and
projects
would
use,
started
off
with
actually
just
EOS
to
govern
over
these
key
admin
functions.
So,
if
you
think
of
a
protocol
identity,
it's
just
having
a
collection
of
public
and
private
functions,
the
public
functions
can
be
called
by.
B
You
know
Ico
process
to
actually
sort
funds
and
have
a
a
safe
way
to
transfer
funds
and
that
as
a
primitive,
has
become
extremely
useful
and
is
used
for
all
sorts
of
things.
Now,
where
it's
not
just.
You
know,
sort
of
a
a
fund.
Storage
and
transfer
system,
but
it's
also
a
way
to
control,
Access
Control
functions
and
do
a
whole
lot
of
things.
B
But
build-based
governance
is
I,
think
the
next
evolution
of
this
process,
but
it
uses
some
of
the
elements.
If,
if
the
you
know
previous
systems,
it
really
integrates
well
with
those,
but
you
kind
of
use
them
in
a
way
that
they
were
supposed
to
be
used
rather
than
how
they're
kind
of
being
used
today.
So
multisigs
are
easy
to
spin
up.
That's
why
they're
used
a
lot.
They
it's
it's
extremely
straightforward.
B
It's
easy
to
kind
of
just
have
a
three
out
of
five
five
out
of
nine
multi-sig,
your
most
practical
purposes,
especially
when
you're
early
kind
of
makes
sense.
But
you
see
a
lot
of
protocols,
not
make
it
clearly,
but
a
lot
of
protocols
that
kind
of
still
use
multi-six
way.
You
know
along
the
development
process,
and
you
know
after
the
boy
after
they
have
product
Market,
fit
and
and
a
lot
of
users,
and
it's
definitely
one
clearly
centralized,
but
also
it's
inflexible.
You
can't
have
a
50
percent
multi-sig.
B
You
know
it's
quite
cumbersome
to
manage.
You
can't
have
sort
of
custom
approval
strategies
where
you
want
a
three
out
of
five
approval
for
this
decision,
but
a
five
out
of
nine
approval
for
this
other
decision,
but,
like
a
you,
know,
25
out
of
50
people
to
approve
you
know
this
decision,
that's
like
a
much
bigger
threshold
and
so
I
think
the
yeah.
B
The
the
sort
of
thing
that
you
know
veto
you
think
about
is
how
you
can
you
know,
use
the
multi-sig
very
well,
but
also
have
the
limitations
set
of
address
the
next
stage
after
the
multi-sig.
It
was
token
voting,
Dows
and-
and
they
are
a
simple
representation
of
a
cap
table
and
it's
it's
been
effective
because
it
kind
of
is
a
simple
one.
Token
one
board,
but
the
problem
is
they're
prone
to
51
attacks.
B
B
Like
you
know,
you're,
seeing
actually
a
news
from
you
know:
VCS
potentially,
like
you
know,
trimming
or
liquidating
these
positions,
so
Mika's
ownership
could
be
more
decentralized,
which
is
really
you
know
interesting
and
exciting,
but
it
but
the
governance
system
isn't
entirely
built.
For
you
know,
a
truly
decentralized
set
of
cap
table
and
I
think
the
other
problems.
B
You
know
that
that
have
emerged
are
you
know,
gendered
apathy
with
token
voting
and
and
less
context
and
decisions,
and
you
know
investors
have
less
context
on
sort
of
decisions
like
granular
decisions
and
protocols
like
how
do
you
update
liquidation
thresholds
or
what
do
you
do
and
the
way
those
problems
are
solved?
Is
you
know,
delegation
which
is
the
apathy
problem
is
solved
when
you
have
the
specific
people
who
spend
a
lot
of
time
in
these
protocols
they
kind
of
really
assess
they
follow
up
they.
B
You
know,
voter
participation,
I
think
is
increased
due
to
delegation
the
you
know,
the
types
of
people
involved
that
I
think
is
expanded.
One
of
the
problems
is
delegation
leads
to
protocols
really
run
by
politicians,
because
delegation
is
a
popularity
contest,
and
so
people
with
a
lot
of
you
know
with
a
huge
following
on
Twitter
or
you
know,
just
have
like
the
the
skills
to
kind
of
maneuver.
B
B
So
what
are
the
functions
that
need
to
be
called
routinely
or-
or
you
know,
in
emergency
cases,
and
who
should
be
the
individuals,
the
groups
or
the
stakeholders
like
token
holders
to
to
make
those
decisions
and
outlining
how
exactly
they
should
do
that.
So,
for
example,
for
a
for
a
routine
like
risk
parameter
upgrade
like
updating,
liquidation
thresholds
or
you
know
updating
kind
of
some
collateralization
parameter.
You
want
that
to
be
done,
but
I.
B
You
know
a
address
core
unit
that
you
just
stopped
out
a
risk
service
provider
and
you
want
that
person
to
be
empowered
to
do
it
as
frequently
as
as
it
needs
to
be
done
in
some
cases.
Maybe
it
needs
to
be
done
multiple
times
a
week,
but
you
need
maybe
a
larger
group,
maybe
it's
the
token
Elders
or
some
some
type
of
larger
group
of
stakeholders
to
be
able
to
veto
or
disapprove
those
decisions
if
they're
made
badly.
B
B
B
You
know
from
all
this
is,
is
about
you
know
the
limitations
of
of
token
voting
and
delegation
and
especially
as
contributed
as
what's
the
optimal
way
to
kind
of
just
get
more
done
and
more
effective
work
done
and
prevent
the
protocol
from
being
say,
captured
by
a
certain
group
of
you
know.
Politicians
or
people
who
are
sub-optimal
for
like
the
effective
sort
of
functioning
of
the
of
the
protocol
and
Salama
is
really
the
framework
we
wish
we
had,
and
so
we've
kind
of
built
it.
B
So
the
three
things
the
Llama
enables
you
to
do.
One
is
encode
Access
Control
policies,
so
we
actually
issue
non-transport
nfts
to
encode
these
Access
Control
policies,
and
they
give
you
permissions
for
what
you
could
do
these
these
nfts
are
non-transport,
so
they're
so
bound.
But
the
cool
thing
is,
you
can
issue
the
nfds
to
an
individual.
You
can
issue
them
to
a
multi-sig
or
to
a
token
voting
down.
B
So
that
way
you
can
actually
have
multiple
stakeholders
kind
of
participate
in
what
the
decision
making
should
be
with
llama.
You
can
create
and
execute
actions,
and
so
this
is
what
we
call
proposal
is
basically
action
in
llama,
and
so
you
know
anything
from
you
know:
transferring
funds
to
updating
protocol
parameters,
anything
that
can
be
represented
on
chain
is
kind
of
what
we
have.
B
You
know
enabled
protocols
to
do
and,
lastly,
manage
funds
programmatically
so
set
rules
for
how
you
manage
funds,
you
could
you
know
the
rules
could
be
automated
or
the
other
rules
could
be.
You
know
done
through
a
specific
Human
Action,
where
you
actually
said.
Like
you
know,
if
it's
let's
say
below
ten
thousand
dollars,
then
the
funds
can
be
approved
with
a
much
lower
approval
threshold,
but
if
it's,
if
it's
a
very
large
amount
to
be
spent,
the
funds
need,
like
you,
know,
a
high
approval
threshold.
B
B
You
know
how
these
sub-dials
function,
but
all
the
routine
stuff
can
happen
at
the
subdot
level,
where
you
know
you
can
kind
of
encode
these
rules
on
a
llama
there's
an
easy
way
to
kind
of
participate
in
these
routine
decisions,
and
you
know
these
sub
accounts
can
be
spent
in
a
useful
way.
But
if
you
say
like
transferring
100
of
the
treasury
of
the
subdot
treasury
to
yourself
like
there's
some
way
that
the
you
know,
maker
governance
can
kind
of
prevent
that
from
happening
yeah.
B
And
so
you
can
issue
these
policies,
maybe
individually
to
individuals,
and
so
you
just
have
you
replace
your
multi-sig
with,
like
you
know,
10
people
that
have
these
policies
and
you
can
have
again
a
custom
approval
strategy
set
up
where
you
know
some
things
require
a
time
lock
of
three
to
five
days.
Some
things
can
be
approved
instantaneously.
B
Some
things
can
be
improved
instantaneously
like
an
emergency,
but
it
requires
a
really
high
approval
threshold
to
be
approved,
instantaneously
and
then
yeah
and
then
and
then
the
last
piece
is
you
can
permission
each
function,
so
you
kind
of
don't
need
to
have
like
a
whole
whole
adoption
of
llama.
You
know
permissioning
every
function,
you
can
actually
start
off
with
just
one
function
and
then
kind
of
you
know
go
on
from
there
where,
if
you
want
to,
you
know,
start
off
with
say
just
seeing
how
to
govern
a
like.
B
A
a
white
listing
function,
a
llama
to
whitelist,
new
vendors
and
service
providers
that
some
protocols
have.
Then
you
can
start
off
with
that
before
you
decide
to
govern
the
pause
function
through
llama
and
and
yeah
and
and
the
way
like
llama's
framework
is,
is
you
know
an
open
source,
public
good
and
so
every
protocol
that
that
uses
llama?
It's
it's
sort
of
they're
using
their
own
instance
of
llama,
and
so
they
have.
B
You
know
the
the
action
is
to
update
liquidation
parameters
and
the
liquidation
threshold
to
90.
You
call
this.
This
update
with
parameter
function
and
the
make
a
main
contract,
and
you
use
like
this
voting
strategy
to
get
it
done,
and
it's
yeah
it's
that
simple
and
you
kind
of
submit
the
action
from
there.
We
kind
of
want
to
take
governance
early
seriously.
We
funded
a
lot
of
times.
B
So
this
is
what
kind
of
the
the
action
page
would
look
like
where
you
just
have
a
list
of
actions
that
have
been
created,
and
so
you
know
this
is
like
the
action
to
update
liquidation
thresholds
through
an
optimistic
strategy
which
is
you
know,
the
service
provider
can
update
it
and
unless
the
rest
of
the
dialogue
group
vetoes
that
it
goes
through,
but
there's
also
these
other
actions
for
transferring
tokens.
This
happens
to
an
optimistic
strategy.
B
Updating
the
stability
fee
happens
through,
say,
a
standard
voting
strategy
where
the
entire
data
should
participate.
Maybe
this
is
like
a
larger
token
transfer.
So
it's
like
it's.
It's
an
emergency
strategy
and
so
yeah
you
just
have
like
these.
You
know
three
components
which
is
what
what
function
you
want
to
call
the
the
Target
contract
and
then
what
strategy
you
want
to
use
yeah.
B
This
is
just
the
the
set
of
action
State
machine
it
it
you
could.
Think
of.
This
is
just
like
how
an
action
goes
from
initial
creation
to
then
approval
and
execution.
So
once
a
Creator
calls
to
create
action
function,
then
an
action
is
active
and
then,
if
the
action
is,
if
the
rules
of
the
action
are
followed-
and
you
know,
the
action
actually
is
grouped,
then
it
gets
approved.
And
finally,
it
gets
queued.
B
Just
like
you
know
normal
proposals
work
and
then
someone
has
to
call
the
execute
function
and
gets
executed.
So
that's
like
the
Happy
case,
pretty
similar
again
to
to
how
standard
governance
Frameworks
work,
but
there's
yeah
there's
a
few
additional
things
where,
where
right
now,
actually
at
least
with
with
Governor
kratoscon
called
the
you
know,
cancel
action
function
which
we
think
is
quite
important
because
there's
a
lot
of
cases
where
I
create
a
creates,
a
proposal.
B
But
then
you
realize
that
there's
there's
an
error
or
something
has
changed
and
the
way
you
say
it's
done.
For
example,
the
navi
governance
is
you
have
to
like?
Remove
your
proposal
power,
that's
kind
of
cumbersome
to
do,
and
so
yeah
I
think
it's
an
important
feature
to
be
able
to
put
the
Creator
to
be
able
to
cancel
in
action
and
and
yeah,
and,
and
so
you
know,
access
actions
can
be
canceled
at
various
stages.
So
that's
a
good
security
thing
where
it
could
be
canceled
by
the
Creator.
B
B
Yeah
sure
so,
I
think
a
good
way
to
think
about.
It
is
yeah.
This
is
a
this
is
what
a
strategy
involves.
So
there's
these
parameters
that
you
need
to
set.
So
this
is
a
this
is
a
standard.
You
know
voting
strategy
again
like
these.
These
names
you
could
Define
it
how
you
want,
but
in
in
this
case
yeah
you
just
enter
your
approval
period
number
of
days
required
to
approve
the
proposal
who
can
approve
this.
B
This
action,
the
number
of
approvals
required
this
can
be
a
number
or
it
can
be
a
percentage.
For
example,
if
it's,
you
know
a
token
running
down
with
10
tens
of
thousands
of
people,
this
could
be
a
percentage
and
then
Force
approval
rewards
which
is
like
who
can,
if
this
person,
if
this
person
with
this
role
has
you
know,
has
the
role
they
can
approve
it
by
default.
So
you
want
to
be
cautious
with
using
that
at
all.
B
And
finally,
you
have
disapproval,
which
is
you
have
a
pedia
to
disapprove,
so
you
can
set
parameters
for
number
of
days
to
disapprove.
Who
can
disapprove?
You
know
what's
the
disapproval
threshold
and
so
an
emergency
strategy
you
can
think
of
is
as
just
the
way
I
think
about
is
like
it's
just
a
strategy
that
has
it's
a
it's.
It's
an
important
action
that
needs
to
be
executed
really
quickly.
So
it's
like
fixing
a
bug
or
you
know,
calling
the
pause
function.
B
B
For
example,
you
keep
one
and
act
on
that,
so
what
an
emergency
strategy
you
know
the
way
you
could
set
the
parameters
is
you
have
a
an
approval
period,
let's
say
three
days,
but
it's
it's
a
variable
approving
approval
period.
So
if
it
gets,
if
you
reach
the
approval
threshold,
it
can
get
executed
approved
immediately
and
you
can
have
a
a
high
approval
threshold.
So
you
can
have,
let's
say,
like
you
know,
30
or
40.
B
Good
policies
and
yeah
again
feel
free
to
interrupt
me
again.
If,
if
there
are
any
questions
suffice
these
are
just
these
non-transport
nfts
that
encode
the
permissions
for
androids,
for
you
know
who
can
create
actions
who
can
submit
approvals
disapprovals
and
once
again
these
policies
they're
really
flexible,
so
they
can
be
issued
to
individuals
like
you
know,
individual
addresses
they
can
be
issued
to
multi-six.
They
can
be
issued
to
contracts
so
token
voting
dial
contracts
would
would
fall
under
this.
B
So
I
think
you
know
the
the
very
simple
version
of
this
is.
Is
you
could
just
have
you
know
you
you
sign
in
with
your
wallet,
you
know,
you
know
you,
you
have
the
the
policy
to
to
you
know
you
have
the
right
to
approve
this
action,
and
so,
if
you
approve
this
option,
you
know
that
your
approval
goes
through.
If
you
don't
have
the
right
to
do
it
like
it's
still,
you
know
it's
still
going
to
be
a
public
there's.
The
community
will
have
public
auditability
of.
B
What's
going
on
so
they'll
be
able
to
see.
You
know
exactly
what
action
is
being
approved
and
disapproved?
Who
has
the
roles
to
do
it,
so
this
I
think
is
quite
important
even
for
your
early
stage
protocols
right,
because
if
you
think
of
a
lot
of
these
learning
stage
protocols,
it
does
make
sense
for
them
to
use
multisigs,
but
but
it's
super
hard
as
a
protocol
user
Community.
B
B
B
You
know
a
way
that
a
few
members,
the
stakeholders
of
a
protocol,
can
can
make
some
decisions,
but
there's
a
way
that
a
community
can
get
involved
in
disapproving
or
beating
those
decisions
if
they,
if
they
kind
of
don't
align
with
with
you
know
what
the
what
the
protocol
was.
You
know
said
it's
going
to
do.
B
So
this
is,
you
know,
accounts
so
accounts.
We,
we
have
kind
of
a
native
accounts
feature
that
is
again
and
you
can
think
of
just
a
contract
that
that
it
holds
it
stores
funds.
You
know
it's
it's
non-custodial
so
and
it's
completely,
you
know
owned
and
controlled
by
a
protocol.
That
is,
you
know,
deploying
their
their
instance
and
and
deploying
their
accounts
and
accounts
can
be
permissioned
the
same
way.
B
Then
you
have
a
really
standard
or
high
approval
threshold
on
this,
where
you
need,
like
you
know
the
standard
token
holder
approval
to
move
funds
from
here,
but
you
actually
have
multiple
sub
accounts,
and
maybe
these
are
like
the
maker
subdials
with
the
endgame
plan
and
those
subdives
just
have
like
these
individuals
or
groups.
Who
can
you
know,
make
these
sort
of
spending
decisions
themselves,
but
maybe,
like
you
know
the
the
make
it
out
token
holders
have
rights
to
some
like
broad
things.
B
C
B
I
spoke
a
little
bit
about
this,
but
I
think
the
the
cool
thing
with
willamas.
You
know,
maker
governance
can
really
outline
the
role
of
token
holders
and
subdas
and
subdivs
can
kind
of
make
decisions
and
routine
matters,
and
they
can
move
quickly
and
they
don't
kind
of
need
this
extreme.
You
know
cumbersome
nature
of
of
standard
token.
A
B
B
C
B
B
Yeah
in
terms
of
make
a
governance
contracts
yeah
like
I
I
described,
but
it
works
pretty
much
across
anything.
So
again
we
can
issue
a
policy
to
yeah.
Like
you
know
the
the
make
a
governance
contract
you
can
issue
it
to
individuals.
You
can
issue
to
subdives.
It
really
doesn't
matter
the
question
on.
B
Can
like
does
every
sub-dial
need
to
adopt
llama
for
it
to
be
useful
or
it
just
can
like
you
know,
wanted
to
do.
Definitely
I
could
I
I.
Think
of
one
or
two
could
could
do
it.
Like
incremental
adoption
is
like
what
we're
designed
for
you?
Don't
you
don't
need
like
we
don't
need
everyone,
everyone
to
adopt
it
for
it
to
be
useful.
A
good
example
is
just
yeah.
B
Let's
just
say
you
start
with
something
like
you
know,
yeah
any
other
service
provider
group
that
that
does
it
and
all
all
it
does
is
there's
a
way
that
you
know,
funds
are
transferred
from.
You
know,
make
it
out
strategy
to
the
subdance
treasury,
and
you
know
most
decisions
are
controlled
by
the
subdial,
but
there's
some,
you
know
critical
decisions.
B
Let's
say
if
you're
moving,
like
you
know,
over
70
of
the
funds
or
something
make
it
as
took
a
notice,
can
can
have
some
some
sort
of
say
in
that.
That's
like
one
light
way
to
do
it,
of
course,
like
the
the
more
interesting
way.
Is
that
if
you,
if
there
are
things
that
are
exact
protocol
functions,
that
are
being
adjusted?
So
if
you're
updating
risk
parameters
and
it's
sort
of
subject
doing
it
and
they
have
the
right
to
do
it,
but
then
again
a
certain
threshold?
B
But
if
they
exceed
that
threshold
then
make
a
doubt
token.
Orders
are
kind
of
involved
in
that
and
so
yeah
we're
definitely
designed
for
incremental
adoption,
where
you
could
initially
just
have
one
group
set
of
tested
out
and
use
it
and
then,
if
it
makes
sense
sort
of
you
know
more
groups
in
common
board
yeah.
If,
if
there's
any
group
that
that
is
interested
or
even
interested
in
like
instead
of
exploring
this,
definitely
let
me
know
we
could
like.
B
B
In
2017,
at
this
Tech
Ranch
conference
and
and
I
have
that
t-shirt
too
yeah.
Maybe
one
day
like
there's
a
there's,
a
better
story,
how
like
llamas,
are
actually
like:
yeah
they're,
hard
involved
in
governance,
day
governance
and
our
governance.
So
I
don't
have
a
great
like
revisionist
history
story,
but
yeah
it's
a
simple
reason
and
yeah.
We
said
llama
started
two
and
a
half
to
you,
know
three
ish
years
back
and
so
there's
a
lot
of
I
I
mean
I.
B
C
B
We
always
want
you
know
a
everyone
to
have
like
a
a
voice
to
express
like
when,
when
things
are
wrong
but
to
to
kind
of
default
like
there's,
you
know
it's
important
for
you
know
for
us
to
build
a
system
where
the
people
who
have
advice
so
not
necessarily
just
the
ones
who
sort
of
gain.
You
know
political
power
in
a
sudden
popularity
contest
and
so
yeah,
that's,
you
know
badly
a
motivation
for
for
building
llama.
C
B
Yeah
I
think
so.
The
previous
work
we've
done
with
protocols
have
all
been
in
the
have
been
more
service
provider,
work
and
so
Lama
is
a
governance
framework.
We've.
Actually
you
know
what
we've
not
like,
AVI
or
others
have
haven't
used
llama
as
the
governance
framework
right
like
they
have
their
own
governance
framework,
and
so
I
mean
more
that
we've
done.
Work
is
the
contributor,
as
a
service
provider,
for
different
projects
and
with
llama
the
governance
frame
and
we've
interacted
with
a
lot
of
government
Frameworks.
B
So
we've
interacted
with
alpha.
Bravo
Avi
governance.
You
know
no
one's
governance
and
others
and
we've
realized
some
of
the
limitations
of
of
these
governance
systems
and
and
therefore,
like
you
know,
build
them
and
kind
of
response
to
that,
and
so
yeah
we're
excited
for
teams
to
use
it
and
and
for
us
to
get
feedback
on
it.
I
think
you
know
it's
always
what
way
we
wrapped
up
our
third
security
audit
and
we're
gonna,
you
know,
deploy
the
framework
soon
and
launch
the
application.
B
B
Yeah,
so
on
on
on
feedback
on
Nama
and
and
kind
of
just
how
how
potential
protocols
and
customers
would
think
about
llama
I
think
the
yeah.
The
feedback
has
been
great,
like
very
positive,
so
we've
been
speaking
to
we're
focused
one
thing.
You
know
it's
important
outline
like
the
types
of
projects
we
want
to
focus
on,
and
so
we're
focused
on
like
our
protocols
and
so
serious,
like
any
smart
contract.
B
Applications
that
are,
you
know,
quote
unquote
like
serious
in
the
sense
that
they're
they're
building
something
that
is
built
to
last.
That
has
some
serious
security
implications.
If
you
know
the
governance
system
fails,
so
not
necessarily
like
the
long
tail
of
on-chain
organizations,
which
some
other,
like
groups
that
are
building
things
in
the
governing
space
are
building
for
that
are
quite
useful
and
cool
like
they
maybe
sell
like
a
long
tail
of
of
applications
and
endows
we're
definitely
building
much
more.
B
For
the
you
know,
the
unit
swaps
and
and
eigenlayers
and
makers
and
optimisms
and
arbitrims
of
the
world,
and
so
anything
from
you
know,
infrastructure
protocols
to
you,
know
other
sort
of
application
protocols
and
default
lending
protocols
and
yeah.
The
feedback
has
been
very
positive
because
I
think
most
times,
teams
kind
of
struggle
with
there's
there's
two
spectrums
where
they
they
added
like
use
a
you
know
earlier
stage
protocols
they
use
the
multi-six.
B
Some
people
carry
on
using
multi-cigarette
for
a
while
after
product
I
could
fit,
and
it's
it's
super
cumbersome
to
you
know,
add
a
lot
more
people
to
the
multisig.
It's
it's
cumbersome
to
have
a
multi
stick
deployed
across
a
lot
of
Chains
It's.
It's
comes
down
to
have
like
these
custom
approval
strategies
or
or
time
blocks
involved
with
the
multi-sig.
B
So
you
know
those
are
problems
with
that
and
then
on
the
other.
On
the
other
hand,
when
you
have
sort
of
the
the
standard
token
voting
dial
with
delegation
system,
it's
just
like
it's
a
hugely
heavy
lift.
First,
like
you
know,
launch
with
the
tokenboarding
now,
because
first
you
have
to
launch
your
token.
So
with
llama.
You,
actually
you
could
choose
to
you
know,
have
a
very
lightweight
governance
system
that
then
progressively
decentralizes,
but
you
need
not
think
about
your
tokenomics
and
governance
yet
and
your
sort
of
token
design.
B
B
But
what
what
we
want
to
see
more
of
is
is
a
really
good
use
of
different
roles
in
governance
in
in
a
way
that
makes
sense,
so
you
know
use
tokenboarding
and
and
token
all
this
to
Governor,
but
things
that
make
sense
that
they
probably
should
govern
over
right,
like
they
should
probably
they.
They
are
good
at
governing
over
say,
like
a
massive
spending
from
the
treasury
or
a
or
like
critical
upgrades
to
the
credit
card,
they're,
probably
not
good
at
governing
over
age
routine.
B
In,
like
you
know,
multiple
times
a
week
or
multiple
times
a
month
like
Risk
parameter,
update
I,
you
know
a
service
provider,
for
example,
should
not
have
like
a
delegated
power
to
then
you
know,
influence
the
approval
of
funds
to
either
them
or
other
service
providers.
We
think
that's
bad
design,
because
it's
like
a
word
incentive
for
service
provider
to
have
delegated
votes
to
then,
like
you
know,
approve
either
their
spending
or
the
spending
of
others.
B
So
we
just
want,
like
a
system
that
uses
all
these
components,
that
if
you
know
we're
always
building
on
the
shoulders
of
giants
and
all
these
components
have
been
built
really
well,
but
they're
kind
of
not
been
used
to
their
for
ability,
because
you
don't
have
this
delineation
of
roles.
That
matter.
B
Let
me
see
I.
Thank
you.
It's
your
name
for
for
your
compliment
and
then
can
Dows
use
these
strategies,
but
tools
like
Snapchat
yeah,
so
Snapchat
is
I,
think
a
great
tool
for
temperature
checks
and,
and
especially
for
you,
you
know
just
like
you
have
like
say
five
options
or
you
know
even
a
few
options
and
it's
a
good
sense
to
get
a
pool
of
of
what
the
community
wants.
B
We're
focused
much
more
on
the
the
finest
serious
final
serious,
like
sort
of
on-chain
version
of
that,
and
we
actually
would
like
more
actions
to
just
come
on
chain
directly.
In
fact,
I
think
in
the
Dallas
that
that
are
interesting
are
the
ones
that
bring
like
almost
all
the
things
on
chain,
and
so,
if
you
see
nouns,
for
example,
they
actually
just
like
default
they're
just
on
chain,
they
actually
don't
even
discuss
much
in
on
a
forum
they
just
like
you
know.
The
proposal
is
on
chain.
B
There's,
like
you
know,
one
day
time,
delay
I
think
there's
like
several
days
to
go
to
the
proposal,
but
it's
on
chain
and
the
discussions
can
happen
everywhere
on
a
discourse
Forum
on
Discord
and
Twitter.
But
the
action
is
kind
of
you
know
beyond
chain
proposal
is
the
is
almost
like
the
source
of
truth.
That
then
leads
to
the
discussion
and
the
debate
and
stuff
elsewhere
that
that
is,
like
you
know,
definitely
it'd
be
cool.
B
To
see
more
of
that,
of
course,
the
guy
doesn't
apply
to
every
type
of
decision,
and
every
type
of
you
know
protocol
where
some
decisions
are
much
more
need,
much
more
like
option
consensus
building
to
get
there
but
I
think,
especially
like
treasury
spending.
Type
decisions
should
just
be.
You
know,
brought
on
chain
one.
B
Yeah
I
mean
I
think
we're
all
kind
of
living
in
the
future
to
some
degree,
like
you
know
at
least
a
few
years
ahead,
or
maybe
five
ten
years
ahead
and
so
yeah
like
you
know,
we
should
try
to
bring
more
things
on
chain
and
and
yeah.
You
know,
there's
gonna
be
problems
with
it,
but
at
least
we
kind
of
discover
the
the
solutions
of
of
bringing
things
excessively
on
chain,
and
how
do
we
solve
for
that?
B
B
Maybe
some
ux
stuff,
but
yeah
I'm
definitely
excited
for
you
know
and,
like
governance,
is
it's
just
like
a
process
for
decision
making
that
you
know
should
be
on
chain
and
yeah
excited
to
see
more
of
that-
and
you
know
it
like
I've
spoken
about
these
like
more
serious
protocol
applications,
but
there's
also
this
like
long
tailor
of
communities
that
I
think
are
interesting
way
for
them.
I
think
governance
is
just
like
an
extension
of
of
social
networks
with
like
real
decision
making
ability.
B
So,
if
you
think
about
say
Constitution
Dao,
if
they
were,
if
they
actually
succeeded-
and
you
know
they
were
more
truly
on
chain
and
they
they
were
able
to
kind
of
govern
over
the
shared
property.
That
would
be
a
really
cool
experiment
in
you
know
how
you
combine
kind
of
social
networks
with
your
decision-making
power
or
like
ownership
over
something
and
yeah
definitely
excited
to
see,
see
more
of
that
too.
B
Yeah,
thank
you
for
having
me
Kesha
and
thanks
everyone
for
the
great
questions
so
I'm
hello,
shreyas
on
on
Twitter
and
and
telegram
you
could.
You
could
reach
me
on
Twitter
and
then
llama's,
just
at
ad
llama
on
Twitter,
so
yeah,
we'll
be
yeah,
we'll
be
sharing
things
more
about
llama
would
not
been,
would
not
publicly
shared
as
as
much
in
the
recent
months.
Babe
we'll
we'll
definitely
do
that
and
would
love
to
get
more
feedback
and
and
input
from
folks
and
yeah.