forcedotcom/sf-skills / skills/sales-agentforce-pipeline-management-configure
sales-agentforce-pipeline-management-configure
VerifiedUse to configure, set up, or repair the Sales Management agent and Agentforce Pipeline Management in a Salesforce org. Automates metadata creation for flows, prompt templates, permission sets, and data source configuration. TRIGGER when: user wants to enable Pipeline Management, configure Sales pipeline features, set up the Sales Management agent for opportunity field updates (including autonomous updates), connect enabled data sources like Einstein Conversation Insights or Einstein Activity Capture, customize opportunity stage descriptions, configure post-meeting suggestions, verify or audit configuration status, fix partially configured orgs, or troubleshoot Pipeline Management metadata issues. DO NOT TRIGGER when: user wants to build a custom agent (use agentforce-generate), configure general Agentforce tracing (use platform-tracing-agentforce-configure), work with non-Sales agents, or enable Einstein Conversation Insights or Einstein Activity Capture from scratch (provisioning is out of scope).
scripts/shared/flow-builder.sh
View on GitHub#!/bin/bash
# shared/flow-builder.sh — build the suggestion flow's field collection from a
# selected set of Opportunity fields.
#
# The base flow template (assets/pipeline_management_flow.flow-meta.xml) ships
# with a hardcoded AddOppFieldsToCollection assignment wiring BOTH NextStep and
# StageName. Historically setup deployed both and then STRIPPED the unwanted one
# (see strip_flow_field in setup-all.sh, still used for the repair path). For a
# greenfield setup we instead build the flow with ONLY the selected fields from
# the start — no deploy-then-strip round trip.
#
# This file only DEFINES a function; it makes no org calls and mutates no files
# when sourced. build_field_collection reads the base template and writes a NEW
# output file; it never edits the template in place.
# build_field_collection <comma-separated-fields> <base-template> <output-file>
#
# Rewrites the AddOppFieldsToCollection assignment so its <assignmentItems>
# contain exactly the given fields (in order, de-duplicated by the caller).
# Uses python3 for a namespace-safe edit — the same requirement as
# strip_flow_field. Returns non-zero (without writing) if python3 is missing,
# the template is unreadable, the assignment/fields are absent, or the write
# fails, so callers can fall back rather than deploy a corrupt/empty flow.
#
# A zero-item AddOppFieldsToCollection is INVALID Flow metadata and fails to
# deploy, so this function refuses to write when the field list is empty.
build_field_collection() {
local fields="$1" base_template="$2" output_file="$3"
if ! command -v python3 >/dev/null 2>&1; then
echo " [FAIL] python3 not available — cannot build flow field collection." >&2
return 1
fi
if [[ ! -f "$base_template" ]]; then
echo " [FAIL] Base flow template not found: $base_template" >&2
return 1
fi
python3 - "$base_template" "$output_file" "$fields" <<'PYEOF'
import sys, xml.etree.ElementTree as ET
base_template, output_file, raw_fields = sys.argv[1], sys.argv[2], sys.argv[3]
# Order-preserving de-dup of trimmed, non-empty field names.
fields = []
for f in raw_fields.split(","):
f = f.strip()
if f and f not in fields:
fields.append(f)
if not fields:
sys.stderr.write(" [FAIL] No fields provided — refusing to write an empty AddOppFieldsToCollection.\n")
sys.exit(2)
ns = "http://soap.sforce.com/2006/04/metadata"
ET.register_namespace("", ns)
try:
tree = ET.parse(base_template)
except Exception as e: # noqa: BLE001
sys.stderr.write(" [FAIL] Could not parse base flow template: %s\n" % e)
sys.exit(3)
root = tree.getroot()
q = lambda t: "{%s}%s" % (ns, t)
target = None
for assign in root.findall(q("assignments")):
name_elem = assign.find(q("name"))
if name_elem is not None and name_elem.text == "AddOppFieldsToCollection":
target = assign
break
if target is None:
sys.stderr.write(" [FAIL] AddOppFieldsToCollection assignment not found in template.\n")
sys.exit(4)
# Drop existing assignmentItems, then rebuild from the selected fields.
for item in list(target.findall(q("assignmentItems"))):
target.remove(item)
# The FlowAssignment schema requires every <assignmentItems> to be contiguous
# and positioned BEFORE <connector>. A plain ET.SubElement(target, ...) appends
# to the end of the element — AFTER <connector> when the template has one — which
# the Metadata API rejects with "Element assignmentItems is duplicated at this
# location". Find the connector's index and insert each new block just before it
# (falling back to append when there is no connector).
conn_idx = next((i for i, c in enumerate(list(target)) if c.tag == q("connector")), None)
insert_at = conn_idx
for field in fields:
item = ET.Element(q("assignmentItems"))
ref = ET.SubElement(item, q("assignToReference"))
ref.text = "OpportunityFields"
op = ET.SubElement(item, q("operator"))
op.text = "Add"
val = ET.SubElement(item, q("value"))
sv = ET.SubElement(val, q("stringValue"))
sv.text = field
if insert_at is None:
target.append(item)
else:
target.insert(insert_at, item)
insert_at += 1
try:
tree.write(output_file, xml_declaration=True, encoding="UTF-8")
except Exception as e: # noqa: BLE001
sys.stderr.write(" [FAIL] Could not write flow output file: %s\n" % e)
sys.exit(5)
sys.exit(0)
PYEOF
}
# strip_flow_field <flow-file> <field> — remove the matching <assignmentItems>
# from the AddOppFieldsToCollection assignment of an EXISTING flow file, in place.
# This is the "fix method" used by the REPAIR path (a flow already deployed with
# fields the admin no longer wants). Greenfield never strips — it builds only the
# selected fields via build_field_collection. Requires python3 for a
# namespace-safe edit; if it is unavailable the field is left wired rather than
# risk corrupting the XML with a regex. Returns 0 if a field was removed, 2 if
# the field was not present (nothing to strip), non-zero on error.
strip_flow_field() {
local flow_file="$1" field="$2"
if ! command -v python3 >/dev/null 2>&1; then
echo " [WARN] python3 not available — cannot strip '$field' from the flow; leaving it wired." >&2
return 1
fi
python3 - "$flow_file" "$field" <<'PYEOF'
import sys, xml.etree.ElementTree as ET
flow_file, field = sys.argv[1], sys.argv[2]
ns = "http://soap.sforce.com/2006/04/metadata"
ET.register_namespace("", ns)
tree = ET.parse(flow_file)
root = tree.getroot()
q = lambda t: "{%s}%s" % (ns, t)
removed = 0
for assign in root.findall(q("assignments")):
name = assign.find(q("name"))
if name is None or name.text != "AddOppFieldsToCollection":
continue
for item in list(assign.findall(q("assignmentItems"))):
sv = item.find("%s/%s" % (q("value"), q("stringValue")))
if sv is not None and sv.text == field:
assign.remove(item)
removed += 1
if removed:
tree.write(flow_file, xml_declaration=True, encoding="UTF-8")
sys.exit(0 if removed else 2)
PYEOF
}
# add_flow_field <flow-file> <field> — idempotently ADD a field's <assignmentItems>
# to the AddOppFieldsToCollection assignment of an EXISTING flow file, in place.
# The symmetric counterpart to strip_flow_field, used by the REPAIR path to wire a
# newly-selected field into an already-deployed flow without a full rebuild (a full
# rebuild would discard the org-specific SalesManagementAgentTemplate constant the
# retrieved flow carries). No-op if the field is already wired. Returns 0 if added,
# 2 if it was already present, non-zero on error.
add_flow_field() {
local flow_file="$1" field="$2"
if ! command -v python3 >/dev/null 2>&1; then
echo " [WARN] python3 not available — cannot add '$field' to the flow." >&2
return 1
fi
python3 - "$flow_file" "$field" <<'PYEOF'
import sys, xml.etree.ElementTree as ET
flow_file, field = sys.argv[1], sys.argv[2]
ns = "http://soap.sforce.com/2006/04/metadata"
ET.register_namespace("", ns)
tree = ET.parse(flow_file)
root = tree.getroot()
q = lambda t: "{%s}%s" % (ns, t)
target = None
for assign in root.findall(q("assignments")):
name = assign.find(q("name"))
if name is not None and name.text == "AddOppFieldsToCollection":
target = assign
break
if target is None:
sys.stderr.write(" [FAIL] AddOppFieldsToCollection assignment not found.\n")
sys.exit(4)
children = list(target)
last_ai_idx = -1
for idx, item in enumerate(children):
if item.tag != q("assignmentItems"):
continue
last_ai_idx = idx
sv = item.find("%s/%s" % (q("value"), q("stringValue")))
if sv is not None and sv.text == field:
sys.exit(2) # already wired
# Build the new assignmentItems block.
item = ET.Element(q("assignmentItems"))
ref = ET.SubElement(item, q("assignToReference")); ref.text = "OpportunityFields"
op = ET.SubElement(item, q("operator")); op.text = "Add"
val = ET.SubElement(item, q("value"))
sv = ET.SubElement(val, q("stringValue")); sv.text = field
# The FlowAssignment schema requires every <assignmentItems> to be contiguous
# and positioned before <connector>. A plain append() lands the new block at
# the end of the element — AFTER <connector> when one exists — which the
# Metadata API rejects with "Element assignmentItems is duplicated at this
# location". Insert immediately after the last existing assignmentItems; if
# there are none, insert just before <connector> (else fall back to append).
if last_ai_idx >= 0:
target.insert(last_ai_idx + 1, item)
else:
conn_idx = next((i for i, c in enumerate(children) if c.tag == q("connector")), None)
if conn_idx is not None:
target.insert(conn_idx, item)
else:
target.append(item)
tree.write(flow_file, xml_declaration=True, encoding="UTF-8")
sys.exit(0)
PYEOF
}
# flow_wired_fields <flow-file> — print the field API names currently wired into
# AddOppFieldsToCollection, one per line (in document order). Used by the repair
# path to diff the deployed flow against the admin's selected set.
flow_wired_fields() {
local flow_file="$1"
command -v python3 >/dev/null 2>&1 || return 1
python3 - "$flow_file" <<'PYEOF'
import sys, xml.etree.ElementTree as ET
ns = "http://soap.sforce.com/2006/04/metadata"
tree = ET.parse(sys.argv[1])
q = lambda t: "{%s}%s" % (ns, t)
for assign in tree.getroot().findall(q("assignments")):
name = assign.find(q("name"))
if name is None or name.text != "AddOppFieldsToCollection":
continue
for item in assign.findall(q("assignmentItems")):
sv = item.find("%s/%s" % (q("value"), q("stringValue")))
if sv is not None and sv.text:
print(sv.text)
PYEOF
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |