mleku/next.orly.dev
1
0
Fork 1
Code Issues 2 Pull Requests Actions Packages Projects Releases Wiki Activity

Add extensive tests and improve policy configuration handling
Some checks failed
Go / build-and-release (push) Has been cancelled
Details

Browse Source 
Introduce comprehensive tests for policy validation logic, including owner and policy admin scenarios. Update `HandlePolicyConfigUpdate` to differentiate permissions for owners and policy admins, enforcing stricter field restrictions and validation flows.
This commit is contained in:
mleku
2025-12-02 07:51:59 +00:00
parent dd8027478c
commit 70944d45df
8 changed files with 1627 additions and 46 deletions
Download Patch File Download Diff File Expand all files Collapse all files

69
app/handle-policy-config.go
View File

@@ -6,7 +6,6 @@ import (
"path/filepath"
"github.com/adrg/xdg"
"lol.mleku.dev/chk"
"lol.mleku.dev/log"
"git.mleku.dev/mleku/nostr/encoders/event"
"git.mleku.dev/mleku/nostr/encoders/filter"
@@ -16,11 +15,20 @@ import (
)
// HandlePolicyConfigUpdate processes kind 12345 policy configuration events.
// Only policy admins can update policy configuration.
// Owners and policy admins can update policy configuration, with different permissions:
//
// OWNERS can:
// - Modify all fields including owners and policy_admins
// - But owners list must remain non-empty (to prevent lockout)
//
// POLICY ADMINS can:
// - Extend rules (add to allow lists, add new kinds, add blacklists)
// - CANNOT modify owners or policy_admins (protected fields)
// - CANNOT reduce owner-granted permissions
//
// Process flow:
// 1. Verify sender is policy admin (from current policy.policy_admins list)
// 2. Parse and validate JSON FIRST (before making any changes)
// 1. Check if sender is owner or policy admin
// 2. Validate JSON with appropriate rules for the sender type
// 3. Pause ALL message processing (lock mutex)
// 4. Reload policy (pause policy engine, update, save, resume)
// 5. Resume message processing (unlock mutex)
@@ -30,24 +38,40 @@ import (
func (l *Listener) HandlePolicyConfigUpdate(ev *event.E) error {
log.I.F("received policy config update from pubkey: %s", hex.Enc(ev.Pubkey))
// 1. Verify sender is policy admin (from current policy.policy_admins list)
// 1. Verify sender is owner or policy admin
if l.policyManager == nil {
return fmt.Errorf("policy system is not enabled")
}
isOwner := l.policyManager.IsOwner(ev.Pubkey)
isAdmin := l.policyManager.IsPolicyAdmin(ev.Pubkey)
if !isAdmin {
log.W.F("policy config update rejected: pubkey %s is not a policy admin", hex.Enc(ev.Pubkey))
return fmt.Errorf("only policy administrators can update policy configuration")
if !isOwner && !isAdmin {
log.W.F("policy config update rejected: pubkey %s is not an owner or policy admin", hex.Enc(ev.Pubkey))
return fmt.Errorf("only owners and policy administrators can update policy configuration")
}
log.I.F("policy admin verified: %s", hex.Enc(ev.Pubkey))
if isOwner {
log.I.F("owner verified: %s", hex.Enc(ev.Pubkey))
} else {
log.I.F("policy admin verified: %s", hex.Enc(ev.Pubkey))
}
// 2. Parse and validate JSON FIRST (before making any changes)
// 2. Parse and validate JSON with appropriate validation rules
policyJSON := []byte(ev.Content)
if err := l.policyManager.ValidateJSON(policyJSON); chk.E(err) {
log.E.F("policy config update validation failed: %v", err)
return fmt.Errorf("invalid policy configuration: %v", err)
var validationErr error
if isOwner {
// Owners can modify all fields, but owners list must be non-empty
validationErr = l.policyManager.ValidateOwnerPolicyUpdate(policyJSON)
} else {
// Policy admins have restrictions: can't modify protected fields, can't reduce permissions
validationErr = l.policyManager.ValidatePolicyAdminUpdate(policyJSON, ev.Pubkey)
}
if validationErr != nil {
log.E.F("policy config update validation failed: %v", validationErr)
return fmt.Errorf("invalid policy configuration: %v", validationErr)
}
log.I.F("policy config validation passed")
@@ -65,12 +89,23 @@ func (l *Listener) HandlePolicyConfigUpdate(ev *event.E) error {
// 4. Reload policy (this will pause policy engine, update, save, and resume)
log.I.F("applying policy configuration update")
if err := l.policyManager.Reload(policyJSON, configPath); chk.E(err) {
log.E.F("policy config update failed: %v", err)
return fmt.Errorf("failed to apply policy configuration: %v", err)
var reloadErr error
if isOwner {
reloadErr = l.policyManager.ReloadAsOwner(policyJSON, configPath)
} else {
reloadErr = l.policyManager.ReloadAsPolicyAdmin(policyJSON, configPath, ev.Pubkey)
}
log.I.F("policy configuration updated successfully by admin: %s", hex.Enc(ev.Pubkey))
if reloadErr != nil {
log.E.F("policy config update failed: %v", reloadErr)
return fmt.Errorf("failed to apply policy configuration: %v", reloadErr)
}
if isOwner {
log.I.F("policy configuration updated successfully by owner: %s", hex.Enc(ev.Pubkey))
} else {
log.I.F("policy configuration updated successfully by policy admin: %s", hex.Enc(ev.Pubkey))
}
// 5. Message processing mutex will be unlocked by defer
return nil