CELER_* env-var prefix cleanup

.github/workflows/e2e.yml

@@ -49,7 +49,7 @@ jobs:
 
     env:
       CGO_ENABLED: '1'
-      CELER_INSECURE_TLS: '1'
+      AGENTPAY_INSECURE_TLS: '1'
 
     steps:
       - name: Checkout

CLAUDE.md

@@ -45,7 +45,7 @@ Wire and admin contracts live in [proto/](proto) and [webapi/proto/](webapi/prot
 
 - **macOS amd64 cgo linker bug.** The local `go1.25.5` toolchain has been observed to fail with duplicate `runtime/cgo` symbols on this codebase (reproduces on trivial `import "C"` programs too — it's a toolchain issue, not a repo regression). Before `go build` / `go test`, export `GOTOOLCHAIN=go1.24.9` for the shell. Documented in [docs/backend-troubleshooting.md](docs/backend-troubleshooting.md).
 - **CGO is required for SQLite-backed builds.** `-storedir` mode fails fast with `sqlite3 requires cgo` if you accidentally build with `CGO_ENABLED=0`. CI sets `CGO_ENABLED=1` explicitly.
-- **`CELER_INSECURE_TLS=1` is normal for localhost.** Inter-OSP and client→OSP localhost dials use the built-in self-signed localhost cert; `test/manual/run_osp.sh` and the e2e harness already set this. Set it yourself when launching binaries directly against `localhost`/`127.0.0.1`.
+- **`AGENTPAY_INSECURE_TLS=1` is normal for localhost.** Inter-OSP and client→OSP localhost dials use the built-in self-signed localhost cert; `test/manual/run_osp.sh` and the e2e harness already set this. Set it yourself when launching binaries directly against `localhost`/`127.0.0.1`.
 - **e2e is slow.** CI gives the default suite 30 minutes (40-minute job cap) and crossnet 30/45. For a local validation loop, prefer the focused single-test runs in [AGENTS.md §Build And Test](AGENTS.md) over `go test ./test/e2e`. On failed runs the harness keeps `/tmp/celer_e2e_*` and prints a `-reuse` path — use it instead of paying the rebuild cost again.
 
 ## Logging convention

config/config.go

@@ -45,10 +45,10 @@ var (
 
 	// Safe-margin knobs are env-var tunable so e2e tests can shrink them. Production
 	// defaults (60s) absorb chain-confirmation slack past a deadline; tests typically
-	// set CELER_*_SAFE_MARGIN_S=5 to keep the timeout-and-sweep flow snappy.
-	WithdrawTimeoutSafeMargin = envUint("CELER_WITHDRAW_SAFE_MARGIN_S", 60) // seconds
-	PaySendTimeoutSafeMargin  = envUint("CELER_PAY_SEND_SAFE_MARGIN_S", 60) // seconds
-	PayRecvTimeoutSafeMargin  = envUint("CELER_PAY_RECV_SAFE_MARGIN_S", 60) // seconds
+	// set AGENTPAY_*_SAFE_MARGIN_S=5 to keep the timeout-and-sweep flow snappy.
+	WithdrawTimeoutSafeMargin = envUint("AGENTPAY_WITHDRAW_SAFE_MARGIN_S", 60) // seconds
+	PaySendTimeoutSafeMargin  = envUint("AGENTPAY_SEND_SAFE_MARGIN_S", 60)     // seconds
+	PayRecvTimeoutSafeMargin  = envUint("AGENTPAY_RECV_SAFE_MARGIN_S", 60)     // seconds
 )
 
 const (

docs/backend-troubleshooting.md

@@ -137,7 +137,7 @@ Checks:
 Known-good local example:
 
 ```bash
-CELER_INSECURE_TLS=1 go run ./server/server.go \
+AGENTPAY_INSECURE_TLS=1 go run ./server/server.go \
   -profile $AGENTPAY_MANUAL_ROOT/profile/o1_profile.json \
   -ks ./testing/env/keystore/osp1.json \
   -port 10001 \
@@ -196,7 +196,7 @@ What these usually mean:
 
 - `celer stream already exists`: the server already has a live or remembered stream for that peer and RPC address.
 - `grpcDial ... failed`: the target host or port is wrong, the peer process is not listening, or TLS/networking is broken.
-- When the target is `localhost` or `127.0.0.1`, a dial timeout can also mean the process is using the built-in self-signed localhost certificate without `CELER_INSECURE_TLS=1` on the dialing side.
+- When the target is `localhost` or `127.0.0.1`, a dial timeout can also mean the process is using the built-in self-signed localhost certificate without `AGENTPAY_INSECURE_TLS=1` on the dialing side.
 - `waitRecvWithTimeout failed`: the transport connected, but the auth handshake did not complete.
 - `peer not online` or `no celer stream`: later traffic depends on a stream that was never established or was dropped.
 

docs/backend-usage.md

@@ -203,7 +203,7 @@ SQLite-backed example:
 ./run_osp.sh 2
 ```
 
-For localhost manual runs, `test/manual/run_osp.sh` defaults `CELER_INSECURE_TLS=1` so inter-OSP dials work with the built-in self-signed localhost certificate.
+For localhost manual runs, `test/manual/run_osp.sh` defaults `AGENTPAY_INSECURE_TLS=1` so inter-OSP dials work with the built-in self-signed localhost certificate.
 
 CockroachDB-backed example:
 
@@ -279,7 +279,7 @@ Optional phase-1 seller-OSP WebAPI listener for a same-host caller:
 -webapigrpc 127.0.0.1:12000
 ```
 
-If this process will dial localhost peers using the built-in localhost certificate, prefix the command with `CELER_INSECURE_TLS=1` unless you are using `test/manual/run_osp.sh`, which already does that for local manual runs.
+If this process will dial localhost peers using the built-in localhost certificate, prefix the command with `AGENTPAY_INSECURE_TLS=1` unless you are using `test/manual/run_osp.sh`, which already does that for local manual runs.
 
 For a CockroachDB-backed node, replace `-storedir` with `-storesql`:
 
@@ -322,13 +322,13 @@ The on-chain contracts use `block.timestamp` (unix seconds) for every challenge
 - rtconfig `min_dispute_timeout` / `max_dispute_timeout` / `max_payment_timeout`
 - per-token rtconfig `min_deadline_delta` / `max_deadline_delta` (open-channel policy)
 - `config.OpenChannelTimeout`, `CooperativeWithdrawTimeout`, `PayResolveTimeout`, `AdminSendTokenTimeout`, `TcbTimeoutSeconds`
-- env-var safe-margin knobs (`CELER_PAY_RECV_SAFE_MARGIN_S`, `CELER_PAY_SEND_SAFE_MARGIN_S`, `CELER_WITHDRAW_SAFE_MARGIN_S`, default `60` each)
+- env-var safe-margin knobs (`AGENTPAY_RECV_SAFE_MARGIN_S`, `AGENTPAY_SEND_SAFE_MARGIN_S`, `AGENTPAY_WITHDRAW_SAFE_MARGIN_S`, default `60` each)
 
 When tuning rtconfig for a new chain, retune in seconds — not blocks. There is no implicit blocks-per-second multiplier in the off-chain code.
 
 #### Test environment overrides
 
-The e2e test harness sets `CELER_*_SAFE_MARGIN_S=5` in `TestMain` so the timeout-and-sweep flow runs in seconds instead of minutes. Production deployments should leave these unset (default `60`).
+The e2e test harness sets `AGENTPAY_*_SAFE_MARGIN_S=5` in `TestMain` so the timeout-and-sweep flow runs in seconds instead of minutes. Production deployments should leave these unset (default `60`).
 
 ## Server Flags That Matter Most
 
@@ -428,7 +428,7 @@ These clients still use the same backend protocol pipeline and storage model des
 
 ## Practical Notes
 
-- The e2e tests set `CELER_INSECURE_TLS=1` so localhost clients can talk to the server's built-in localhost certificate without CA setup.
+- The e2e tests set `AGENTPAY_INSECURE_TLS=1` so localhost clients can talk to the server's built-in localhost certificate without CA setup.
 - OSP routing behavior only becomes meaningful after the OSP is registered in the on-chain `RouterRegistry`.
 - The server starts periodic OSP cleanup that clears expired or on-chain-resolved payments with peer OSPs.
 - `rtconfig` is operationally important. Payment timeout, refill, and deposit behavior are not hardcoded solely in Go constants.

test/e2e/e2e_setup_test.go

@@ -30,13 +30,13 @@ func TestMain(m *testing.M) {
 	log.EnableColor()
 	// Allow client dials to use insecure transport for localhost during e2e,
 	// avoiding CA mismatches with the server's self-signed localhost cert.
-	os.Setenv("CELER_INSECURE_TLS", "1")
+	os.Setenv("AGENTPAY_INSECURE_TLS", "1")
 	// Shrink the chain-confirmation slack past pay/withdraw deadlines so the
 	// timeout-and-sweep e2e flow runs in seconds instead of minutes. Production
 	// defaults (60s) are restored automatically when these env vars are unset.
-	os.Setenv("CELER_PAY_RECV_SAFE_MARGIN_S", "5")
-	os.Setenv("CELER_PAY_SEND_SAFE_MARGIN_S", "5")
-	os.Setenv("CELER_WITHDRAW_SAFE_MARGIN_S", "5")
+	os.Setenv("AGENTPAY_RECV_SAFE_MARGIN_S", "5")
+	os.Setenv("AGENTPAY_SEND_SAFE_MARGIN_S", "5")
+	os.Setenv("AGENTPAY_WITHDRAW_SAFE_MARGIN_S", "5")
 	// Ensure DEBUG and above from app are visible in test output by default
 	log.SetLevelByName("debug")
 	if *reuse != "" {

test/e2e/send_pay_timeout.go

@@ -152,7 +152,7 @@ func sendPayTimeout(t *testing.T, tokenType entity.TokenType, tokenAddr string)
 		return
 	}
 
-	// Wait past the pay deadline + receiver-side safe margin (CELER_PAY_RECV_SAFE_MARGIN_S=5
+	// Wait past the pay deadline + receiver-side safe margin (AGENTPAY_RECV_SAFE_MARGIN_S=5
 	// in TestMain) so SettleExpiredPays sees the pay as expired.
 	err = c1.WaitUntilDeadline(payTime + timeout + 10)
 	if err != nil {

test/manual/README.md

@@ -38,7 +38,7 @@ Then do the same for OSP2, run **`./osp-cli -profile $AGENTPAY_MANUAL_ROOT/profi
 
 Run **`./run_osp.sh 1`** and **`./run_osp.sh 2`** in two new terminals respectively to start OSP1 and OSP2. OSP data store is created at `$AGENTPAY_MANUAL_ROOT/store/[ospAddr]`.
 
-For local manual runs, `run_osp.sh` defaults `CELER_INSECURE_TLS=1` so inter-OSP dials work with the built-in self-signed localhost certificate. If you start the servers directly instead of using this script, set that env var yourself.
+For local manual runs, `run_osp.sh` defaults `AGENTPAY_INSECURE_TLS=1` so inter-OSP dials work with the built-in self-signed localhost certificate. If you start the servers directly instead of using this script, set that env var yourself.
 
 ### Option 2: use CockroachDB as storage backend (higher performance)
 

test/manual/run_osp.sh

@@ -1,11 +1,11 @@
 #!/bin/sh
 
 MANUAL_ROOT="${AGENTPAY_MANUAL_ROOT:-/tmp/celer_manual_test}"
-LOCAL_INSECURE_TLS="${CELER_INSECURE_TLS:-1}"
+LOCAL_INSECURE_TLS="${AGENTPAY_INSECURE_TLS:-1}"
 
 run_osp_1() {
   echo "run OSP 1"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o1_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp1.json" \
     -port 10001 \
@@ -21,7 +21,7 @@ run_osp_1() {
 
 run_osp_1_crdb() {
   echo "run OSP 1 w/ cockroach db"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o1_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp1.json" \
     -port 10001 \
@@ -37,7 +37,7 @@ run_osp_1_crdb() {
 
 run_osp_2() {
   echo "run OSP 2"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o2_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp2.json" \
     -port 10002 \
@@ -53,7 +53,7 @@ run_osp_2() {
 
 run_osp_2_crdb() {
   echo "run OSP 2 w/ cockroach db"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o2_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp2.json" \
     -port 10002 \
@@ -69,7 +69,7 @@ run_osp_2_crdb() {
 
 run_osp_3() {
   echo "run OSP 3"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o3_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp3.json" \
     -port 10003 \
@@ -85,7 +85,7 @@ run_osp_3() {
 
 run_osp_3_crdb() {
   echo "run OSP 3 w/ cockroach db"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o3_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp3.json" \
     -port 10003 \
@@ -101,7 +101,7 @@ run_osp_3_crdb() {
 
 run_osp_4() {
   echo "run OSP 4"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o4_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp4.json" \
     -port 10004 \
@@ -117,7 +117,7 @@ run_osp_4() {
 
 run_osp_4_crdb() {
   echo "run OSP 4 w/ cockroach db"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o4_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp4.json" \
     -port 10004 \
@@ -133,7 +133,7 @@ run_osp_4_crdb() {
 
 run_osp_5() {
   echo "run OSP 5"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o5_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp5.json" \
     -port 10005 \
@@ -149,7 +149,7 @@ run_osp_5() {
 
 run_osp_5_crdb() {
   echo "run OSP 5 w/ cockroach db"
-  CELER_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
+  AGENTPAY_INSECURE_TLS="${LOCAL_INSECURE_TLS}" go run "${AGENTPAY}/server/server.go" \
     -profile "${MANUAL_ROOT}/profile/o5_profile.json" \
     -ks "${AGENTPAY}/testing/env/keystore/osp5.json" \
     -port 10005 \

utils/utils.go

@@ -115,12 +115,12 @@ func GetClientTlsOption() grpc.DialOption {
 }
 
 func IsPermissiveClientTLS() bool {
-	v := os.Getenv("CELER_INSECURE_TLS")
+	v := os.Getenv("AGENTPAY_INSECURE_TLS")
 	return v == "1" || strings.EqualFold(v, "true")
 }
 
 // GetClientTlsOptionPermissive returns insecure transport credentials when the
-// environment variable CELER_INSECURE_TLS is set ("1"/"true"). This is useful
+// environment variable AGENTPAY_INSECURE_TLS is set ("1"/"true"). This is useful
 // for local e2e tests where the server uses a self-signed localhost certificate
 // that may not chain to CAs available to the client.
 func GetClientTlsOptionPermissive() grpc.DialOption {
@@ -149,7 +149,7 @@ func WrapLocalTLSDialError(target string, err error) error {
 	if err == nil || IsPermissiveClientTLS() || !IsLoopbackTarget(target) {
 		return err
 	}
-	return fmt.Errorf("%w; local TLS hint: %s uses the built-in self-signed localhost certificate, set CELER_INSECURE_TLS=1 on the dialing process or configure a trusted cert via -tlscert/-tlskey", err, target)
+	return fmt.Errorf("%w; local TLS hint: %s uses the built-in self-signed localhost certificate, set AGENTPAY_INSECURE_TLS=1 on the dialing process or configure a trusted cert via -tlscert/-tlskey", err, target)
 }
 
 // GetClientTlsConfig returns tls.Config with system and celerCA, for https interaction